From david.huen at ntlworld.com  Sat Jan  3 09:32:53 2004
From: david.huen at ntlworld.com (David Huen)
Date: Sat Jan  3 09:40:08 2004
Subject: [Biojava-dev] Current Alphabet design and an unintended consequence
	for training
Message-ID: <200401031432.53811.david.huen@ntlworld.com>

The current Alphabet system uses the BasisSymbol to represent both ambiguity 
symbols and symbols from cross product alphabets.  This casues unintended 
consequences for training algorithms.

The DistributionTrainerContext has an addCount(Distribution dist, Symbol 
sym, double times) method.  When using cross product alphabets, it works 
flawlessly when it encounters AtomicSymbols from the cross-product alphabet 
(and these are also BasisSymbols).  In the design of the 
DistributionTrainer interface, the equivalent method addCount(Distribution 
dist, AtomicSymbol sym, double times) accepts only an AtomicSymbol, which 
is reasonable.

However, when training two-head distributions, it is not implausible for the 
DistributionTrainerContext.addCount() to receive Symbols that are not 
AtomicSymbols.  The most common by far would be symbols emitted by gap 
states of form e.g (gap cytosine).  The current implementation of the 
addCount method assumes that non-atomic symbols are ambiguity symbols and 
attempts to deal with them in that manner.  Evidently it fails in the above 
case, indeed, it fails silently.  This problem currently prevents the 
training of PairDistributions in which one component Distribution is a 
GapDistribution.

There appears to be no easy way of fixing this problem at the level of   
DistributionTrainerContext.  It is formally possible that the BasisSymbol 
received by addCount is truly an ambiguity symbol containing a number of 
symbols from the cross-product alphabet of the two-head HMM model.  It is 
also possible that the BasisSymbol represents a single symbol comprising 
ambiguity symbol(s) from one or both alphabets that form the cross product 
alphabet.  The two are evidently not equivalent and have to be dealt with 
differently.  And resolving which it is is potentially computationally 
costly for an operation that is repeated very many times during training.

Even if this ambiguity could be resolved at the level of 
DistributionTrainerContext and you knew the symbol to be one of type (gap 
<something else>), that symbol cannot be passed to a DistributionTrainer 
that may be capable of dealing with it as the addCount method in that 
interface accepts only atomic symbols which something like (gap guanine) is 
not.

Interim solutions could be:-
1) change the DistributionTrainer.addCount()  to accept non-atomic symbols.  
DistributionTrainerContext's addCount method will leave it to the 
distribution trainers to sort out what to do with non-atomic symbols 
themselves.  
OR
2) add a ExtendedDistributionTrainer interface with one method addCount that 
can accept non-atomic symbols.  DistributionTrainerContext's addCount 
method will check whether the symbol it receives is atomic.  If it is, it 
will use the standard DistributionTrainer.addCount().  If not, it will 
determine if the trainer for that distribution implements the 
ExtendedDistributionTrainer interface and if so, call that interface's 
addCount method to leave it to deal with the symbol.  If not, it will 
assume that the symbol is an ambiguity symbol and deal with it in the 
manner it does now.

(2) is probably less disruptive to existing code and interfaces.  It may be 
that the DistributionTrainer is a better place to deal with non-atomic 
symbols than DistributionTrainerContext since that the DT knows more about 
the internals of that Distribution and what it can/should handle while the 
DTC has to of necessity implement a one-size-fits-all approach.

At Biojava 2, it may be worthwhile to revisit the Alphabet design and 
explicitly distinguish ambiguity symbols and BasisSymbols on the level that 
the former is a Set of symbols, while the latter is a List of Symbols.

Regards,
David Huen





From markjschreiber at hotmail.com  Sat Jan  3 04:13:21 2004
From: markjschreiber at hotmail.com (Mark Schreiber)
Date: Sat Jan  3 12:17:31 2004
Subject: [Biojava-dev] DP memory leak
Message-ID: <BAY7-F3mCClDBdMFqxL00002a67@hotmail.com>

An HTML attachment was scrubbed...
URL: http://portal.open-bio.org/pipermail/biojava-dev/attachments/20040103/99862e81/attachment.htm
From markjschreiber at hotmail.com  Sun Jan  4 08:30:48 2004
From: markjschreiber at hotmail.com (mark schreiber)
Date: Sun Jan  4 08:37:19 2004
Subject: [Biojava-dev] Current Alphabet design and an unintended
	consequencefor training
In-Reply-To: <200401031432.53811.david.huen@ntlworld.com>
Message-ID: <BAY7-DAV38KvT7sGJpK000210f6@hotmail.com>

David,

I think you may be able to use solution 1. As Symbol is a superclass
(interface) of AtomicSymbol you won't be breaking any API.

- Mark


> -----Original Message-----
> From: biojava-dev-bounces@portal.open-bio.org 
> [mailto:biojava-dev-bounces@portal.open-bio.org] On Behalf Of 
> David Huen
> Sent: Saturday, 3 January 2004 10:33 p.m.
> To: biojava-dev@biojava.org
> Subject: [Biojava-dev] Current Alphabet design and an 
> unintended consequencefor training
> 
> The current Alphabet system uses the BasisSymbol to represent 
> both ambiguity symbols and symbols from cross product 
> alphabets.  This casues unintended consequences for training 
> algorithms.
> 
> The DistributionTrainerContext has an addCount(Distribution 
> dist, Symbol sym, double times) method.  When using cross 
> product alphabets, it works flawlessly when it encounters 
> AtomicSymbols from the cross-product alphabet (and these are 
> also BasisSymbols).  In the design of the DistributionTrainer 
> interface, the equivalent method addCount(Distribution dist, 
> AtomicSymbol sym, double times) accepts only an AtomicSymbol, 
> which is reasonable.
> 
> However, when training two-head distributions, it is not 
> implausible for the
> DistributionTrainerContext.addCount() to receive Symbols that 
> are not AtomicSymbols.  The most common by far would be 
> symbols emitted by gap states of form e.g (gap cytosine).  
> The current implementation of the addCount method assumes 
> that non-atomic symbols are ambiguity symbols and attempts to 
> deal with them in that manner.  Evidently it fails in the 
> above case, indeed, it fails silently.  This problem 
> currently prevents the training of PairDistributions in which 
> one component Distribution is a GapDistribution.
> 
> There appears to be no easy way of fixing this problem at the 
> level of   
> DistributionTrainerContext.  It is formally possible that the 
> BasisSymbol received by addCount is truly an ambiguity symbol 
> containing a number of symbols from the cross-product 
> alphabet of the two-head HMM model.  It is also possible that 
> the BasisSymbol represents a single symbol comprising 
> ambiguity symbol(s) from one or both alphabets that form the 
> cross product alphabet.  The two are evidently not equivalent 
> and have to be dealt with differently.  And resolving which 
> it is is potentially computationally costly for an operation 
> that is repeated very many times during training.
> 
> Even if this ambiguity could be resolved at the level of 
> DistributionTrainerContext and you knew the symbol to be one 
> of type (gap <something else>), that symbol cannot be passed 
> to a DistributionTrainer that may be capable of dealing with 
> it as the addCount method in that interface accepts only 
> atomic symbols which something like (gap guanine) is not.
> 
> Interim solutions could be:-
> 1) change the DistributionTrainer.addCount()  to accept 
> non-atomic symbols.  
> DistributionTrainerContext's addCount method will leave it to 
> the distribution trainers to sort out what to do with 
> non-atomic symbols themselves.  
> OR
> 2) add a ExtendedDistributionTrainer interface with one 
> method addCount that can accept non-atomic symbols.  
> DistributionTrainerContext's addCount method will check 
> whether the symbol it receives is atomic.  If it is, it will 
> use the standard DistributionTrainer.addCount().  If not, it 
> will determine if the trainer for that distribution 
> implements the ExtendedDistributionTrainer interface and if 
> so, call that interface's addCount method to leave it to deal 
> with the symbol.  If not, it will assume that the symbol is 
> an ambiguity symbol and deal with it in the manner it does now.
> 
> (2) is probably less disruptive to existing code and 
> interfaces.  It may be that the DistributionTrainer is a 
> better place to deal with non-atomic symbols than 
> DistributionTrainerContext since that the DT knows more about 
> the internals of that Distribution and what it can/should 
> handle while the DTC has to of necessity implement a 
> one-size-fits-all approach.
> 
> At Biojava 2, it may be worthwhile to revisit the Alphabet 
> design and explicitly distinguish ambiguity symbols and 
> BasisSymbols on the level that the former is a Set of 
> symbols, while the latter is a List of Symbols.
> 
> Regards,
> David Huen
> 
> 
> 
> 
> 
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
> 
From david.huen at ntlworld.com  Sun Jan  4 16:46:04 2004
From: david.huen at ntlworld.com (David Huen)
Date: Sun Jan  4 16:53:14 2004
Subject: [Biojava-dev] In which the Multihead DP traverses MagicalState
	repeatedly
Message-ID: <200401042146.04716.david.huen@ntlworld.com>

With some combinations of parameters (rather poor ones for the sequences to 
aligned by Viterbi), our twohead DP implementation can generate a StatePath 
that traverses the magicalState on multiple occasions.

After investigating it at length, I think it comes about comes about because 
the EmissionCache treats the MagicalState as an EmissionState with advance 
(0, 0) and a finite probability.  However, traversal through the 
MagicalState should be explicitly forbidden in the Viterbi statepath.

I have implemented fixes in both LightPairDPCursor and EmissionCache to set 
the probability score of the MagicalState to -infinity at all points in the 
DP matrix other than at the origin and bottom right corner.  This seems to 
have fixed this problem within my dataset.

I have also retrofitted an equivalent fix into AbstractMatrixPairDPCursor 
but I am insufficiently familiar with that code to be certain of it being 
correct.

I don't know if there are consequences for DPCompiler.

Anyway, I have committed the fixes.

Incidentally, if you have been running 2-head DPs with the current code and 
do not usually examine the generated StatePaths, you may wish to repeat 
some of those runs and compare the results.  The ability to traverse the 
MagicalState does create amazing opportunities for the model to go where no 
model has ever been before :-)

Regards,
David Huen
From david.huen at ntlworld.com  Sun Jan  4 17:13:36 2004
From: david.huen at ntlworld.com (David Huen)
Date: Sun Jan  4 17:20:43 2004
Subject: [Biojava-dev] Extensions to DP framework to permit 2-head training
Message-ID: <200401042213.36593.david.huen@ntlworld.com>

I have written a working Viterbi trainer that is capable of 1 and 2 head 
training and hope to commit to CVS. The current API does not permit 2-head 
training and will need changes to accomodate this.

I propose the following changes to permit it:-

a) introduction of a TrainingSet interface: In 2-head training, pairs of 
sequences need to be available.  I have generalised it into a means of 
supplying n sequences per case so we can deal with n-head training when 20 
gazillion Hz processors with 5 bazillion byte RAM become available.

public interface TrainingSet
{
    public interface Iterator
    {
        /**
         * get next group of sequences to train the model on.
         */
        public Sequence[] next();

        /**
         * any further training sequence groups?
         */
        public boolean hasNext();
    }

    /**
     * get an iterator for the cases supplied by this TrainingSet.
     */
    public Iterator getCases();
}

b) changes to TrainingAlgorithm interface:-

The current train method takes a SequenceDB which only works for 1-head 
training.  I propose a further method that takes a TrainingSet and 
deprecating the current method.

This change will break code that derives from AbstractTrainer.  But I think 
I can cursorily patch up the AbstractTrainer and current BaumWelch code to 
add the new call.  I do not propose extending the BW code to handling 2-D 
training at this stage.

c) changes to AbstractTrainer class:-
AbstractTrainer supplies a 

protected abstract double singleSequenceIteration(ModelTrainer trainer, 
SymbolList symList)

I propose changing it to:-
protected abstract double singleSequenceIteration(ModelTrainer trainer, 
SymbolList [] symList)

Perhaps to avoid breaking too much, I should create a NewAbstractTrainer 
class with the new method and derive my ViterbiTrainer from it instead

Comments are requested.

Regards,
David Huen

From benjamins at Biomax.de  Mon Jan  5 08:15:02 2004
From: benjamins at Biomax.de (Benjamin Schuster-Boeckler)
Date: Mon Jan  5 08:22:13 2004
Subject: [Biojava-dev] Why isn't SimpleAnnotation in org.biojava.bio.impl ?
Message-ID: <3FF96356.5040404@Biomax.de>

I wondered because generally I thought all implementing classes should 
go into a dedicated impl directory... I guess this is only for backward 
compatibility then?!

Greetings,

Benjamin Schuster-B?ckler


From thomas at derkholm.net  Mon Jan  5 09:18:04 2004
From: thomas at derkholm.net (Thomas Down)
Date: Mon Jan  5 08:35:08 2004
Subject: [Biojava-dev] Why isn't SimpleAnnotation in org.biojava.bio.impl ?
In-Reply-To: <3FF96356.5040404@Biomax.de>
References: <3FF96356.5040404@Biomax.de>
Message-ID: <20040105141804.GA22554@firechild>

Once upon a time, Benjamin Schuster-Boeckler wrote:
> I wondered because generally I thought all implementing classes should 
> go into a dedicated impl directory... I guess this is only for backward 
> compatibility then?!

We like the idea of keeping implementations separate from interfaces
using impl packages, but unfortunately we didn't think of this when
we first started BioJava.  The Annotation interface (and SimpleAnnotation)
go right back to pre-1.0 versions of BioJava.

There has been a certain amount of discussion about doing some
major redesign and refactoring to produce a "BioJava 2.0" -- that's
probably the best time to make an interface/implementation split
universal.

     Thomas.
From thomas at derkholm.net  Tue Jan  6 08:55:05 2004
From: thomas at derkholm.net (Thomas Down)
Date: Tue Jan  6 08:12:08 2004
Subject: [Biojava-dev] DP memory leak
In-Reply-To: <3FE6DE38.1090000@yahoo.co.uk>
References: <3FE1C160.2040906@nrc-cnrc.gc.ca> <3FE32525.5000002@sanger.ac.uk>
	<3FE6DE38.1090000@yahoo.co.uk>
Message-ID: <20040106135504.GD23174@firechild>

Did this problem ever get fixed?  I've got a DP application
leaking pretty badly right now.

If the answer is no, would anyone object to me implementing
the weak-reference solution? (which might require a little
bit of magic so that the `well known' alphabets don't get
unloaded immediately after BJ initializes.  *sigh*).

     Thomas.

Once upon a time, Matthew Pocock wrote:
> Hi,
> 
> Lachlan James Coin wrote:
> 
> >Hi,
> >
> >I came across a bit of a memory leak while using the dp code.  The 
> >problem is that AlphabetManager keeps a reference to every alphabet 
> >created, including StateAlphabetcreated when using the dp package.  I 
> >had extended the EmissionState class in such a way that the instances 
> >were of a non-trivial size.  I am also creating (one after the other) 
> >a lot of different state alphabets.  The result was that I ran out of 
> >memory farily quickly.
> 
> Hehe. That's an unforseen use-case bug.
> 
> >
> >One possible fix I thought of was to add "deregisterAlphabet(name)" 
> >method to AlphabetManager, and to call this from the finalize mehod of 
> >SimpleMarkovModel, so that when the garbage collector thinks it can 
> >clear the markov model the statealphabet is removed from the 
> >AlphabetManager registry and can then also be garbage collected.   
> >Also, it would be necessary to make alphabetToIndex a WeakHashMap.
> 
> This may work - but it adds API.
> 
> >
> >Any other suggestions?  Can I make these changes?
> 
> How transparent can we make this? How easy would it be for alphabets to 
> be dropped automatically when not needed any more? I guess we don't want 
> to drop DNA & friends :)
> 
> >
> >Thanks,
> >
> >Lachlan
> >
> Matthew
> 
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
From lc1 at sanger.ac.uk  Tue Jan  6 08:13:27 2004
From: lc1 at sanger.ac.uk (Lachlan James Coin)
Date: Tue Jan  6 08:20:31 2004
Subject: [Biojava-dev] DP memory leak
In-Reply-To: <20040106135504.GD23174@firechild>
References: <3FE1C160.2040906@nrc-cnrc.gc.ca> <3FE32525.5000002@sanger.ac.uk>
	<3FE6DE38.1090000@yahoo.co.uk> <20040106135504.GD23174@firechild>
Message-ID: <3FFAB477.4020501@sanger.ac.uk>

I wasn't sure if there was enough consensus to commit my changes, so I 
haven't fixed anything yet:
 - adding a deregisterAlphabet(name) method to AlphabetManager and 
calling this from finalize in SimpleMarkovModel
- making alphabetToIndex a WeakHashMap  (I'm not 100% sure that this is 
OK with regard to standard alphabets)


Lachlan




Thomas Down wrote:

>Did this problem ever get fixed?  I've got a DP application
>leaking pretty badly right now.
>
>If the answer is no, would anyone object to me implementing
>the weak-reference solution? (which might require a little
>bit of magic so that the `well known' alphabets don't get
>unloaded immediately after BJ initializes.  *sigh*).
>
>     Thomas.
>
>Once upon a time, Matthew Pocock wrote:
>  
>
>>Hi,
>>
>>Lachlan James Coin wrote:
>>
>>    
>>
>>>Hi,
>>>
>>>I came across a bit of a memory leak while using the dp code.  The 
>>>problem is that AlphabetManager keeps a reference to every alphabet 
>>>created, including StateAlphabetcreated when using the dp package.  I 
>>>had extended the EmissionState class in such a way that the instances 
>>>were of a non-trivial size.  I am also creating (one after the other) 
>>>a lot of different state alphabets.  The result was that I ran out of 
>>>memory farily quickly.
>>>      
>>>
>>Hehe. That's an unforseen use-case bug.
>>
>>    
>>
>>>One possible fix I thought of was to add "deregisterAlphabet(name)" 
>>>method to AlphabetManager, and to call this from the finalize mehod of 
>>>SimpleMarkovModel, so that when the garbage collector thinks it can 
>>>clear the markov model the statealphabet is removed from the 
>>>AlphabetManager registry and can then also be garbage collected.   
>>>Also, it would be necessary to make alphabetToIndex a WeakHashMap.
>>>      
>>>
>>This may work - but it adds API.
>>
>>    
>>
>>>Any other suggestions?  Can I make these changes?
>>>      
>>>
>>How transparent can we make this? How easy would it be for alphabets to 
>>be dropped automatically when not needed any more? I guess we don't want 
>>to drop DNA & friends :)
>>
>>    
>>
>>>Thanks,
>>>
>>>Lachlan
>>>
>>>      
>>>
>>Matthew
>>
>>_______________________________________________
>>biojava-dev mailing list
>>biojava-dev@biojava.org
>>http://biojava.org/mailman/listinfo/biojava-dev
>>    
>>
>
>  
>


From thomas at derkholm.net  Tue Jan  6 09:36:49 2004
From: thomas at derkholm.net (Thomas Down)
Date: Tue Jan  6 08:53:53 2004
Subject: [Biojava-dev] DP memory leak
In-Reply-To: <3FFAB477.4020501@sanger.ac.uk>
References: <3FE1C160.2040906@nrc-cnrc.gc.ca> <3FE32525.5000002@sanger.ac.uk>
	<3FE6DE38.1090000@yahoo.co.uk> <20040106135504.GD23174@firechild>
	<3FFAB477.4020501@sanger.ac.uk>
Message-ID: <20040106143649.GE23174@firechild>

Once upon a time, Lachlan James Coin wrote:
> I wasn't sure if there was enough consensus to commit my changes, so I 
> haven't fixed anything yet:
> - adding a deregisterAlphabet(name) method to AlphabetManager and 
> calling this from finalize in SimpleMarkovModel

I'm not a big finalize fan, but I guess this makes sense.

> - making alphabetToIndex a WeakHashMap  (I'm not 100% sure that this is 
> OK with regard to standard alphabets)

Does this actually help?  The AlphabetIndex implementations keep
hard references to their corresponding Alphabets.  I thought
WeakHashMap still kept hard references to values until their keys
are enqueued, so I don't see how an Alphabet can ever be collected.

Or am I missing something?

     Thomas.
From lc1 at sanger.ac.uk  Tue Jan  6 08:54:57 2004
From: lc1 at sanger.ac.uk (Lachlan James Coin)
Date: Tue Jan  6 09:02:00 2004
Subject: [Biojava-dev] DP memory leak
In-Reply-To: <20040106143649.GE23174@firechild>
References: <3FE1C160.2040906@nrc-cnrc.gc.ca> <3FE32525.5000002@sanger.ac.uk>
	<3FE6DE38.1090000@yahoo.co.uk> <20040106135504.GD23174@firechild>
	<3FFAB477.4020501@sanger.ac.uk> <20040106143649.GE23174@firechild>
Message-ID: <3FFABE31.7000706@sanger.ac.uk>

OK - good point.  I don't really understand the AlphabetManager code 
that well.  

Can we put something in the deregisterAlphabet method that will take 
care of this?

Lachlan


Thomas Down wrote:

>Once upon a time, Lachlan James Coin wrote:
>  
>
>>I wasn't sure if there was enough consensus to commit my changes, so I 
>>haven't fixed anything yet:
>>- adding a deregisterAlphabet(name) method to AlphabetManager and 
>>calling this from finalize in SimpleMarkovModel
>>    
>>
>
>I'm not a big finalize fan, but I guess this makes sense.
>
>  
>
>>- making alphabetToIndex a WeakHashMap  (I'm not 100% sure that this is 
>>OK with regard to standard alphabets)
>>    
>>
>
>Does this actually help?  The AlphabetIndex implementations keep
>hard references to their corresponding Alphabets.  I thought
>WeakHashMap still kept hard references to values until their keys
>are enqueued, so I don't see how an Alphabet can ever be collected.
>
>Or am I missing something?
>
>     Thomas.
>
>  
>


From thomas at derkholm.net  Tue Jan  6 10:29:57 2004
From: thomas at derkholm.net (Thomas Down)
Date: Tue Jan  6 09:47:00 2004
Subject: [Biojava-dev] DP memory leak
In-Reply-To: <20040106143649.GE23174@firechild>
References: <3FE1C160.2040906@nrc-cnrc.gc.ca> <3FE32525.5000002@sanger.ac.uk>
	<3FE6DE38.1090000@yahoo.co.uk> <20040106135504.GD23174@firechild>
	<3FFAB477.4020501@sanger.ac.uk> <20040106143649.GE23174@firechild>
Message-ID: <20040106152957.GB23331@firechild>

Okay, I've got this working now by making LinearAlphabetIndex
keep a WeakReference to its Alphabet.

    Thomas.

Once upon a time, Thomas Down wrote:
> Once upon a time, Lachlan James Coin wrote:
> > I wasn't sure if there was enough consensus to commit my changes, so I 
> > haven't fixed anything yet:
> > - adding a deregisterAlphabet(name) method to AlphabetManager and 
> > calling this from finalize in SimpleMarkovModel
> 
> I'm not a big finalize fan, but I guess this makes sense.
> 
> > - making alphabetToIndex a WeakHashMap  (I'm not 100% sure that this is 
> > OK with regard to standard alphabets)
> 
> Does this actually help?  The AlphabetIndex implementations keep
> hard references to their corresponding Alphabets.  I thought
> WeakHashMap still kept hard references to values until their keys
> are enqueued, so I don't see how an Alphabet can ever be collected.
> 
> Or am I missing something?
> 
>      Thomas.
From mark.schreiber at group.novartis.com  Tue Jan  6 20:19:53 2004
From: mark.schreiber at group.novartis.com (mark.schreiber@group.novartis.com)
Date: Tue Jan  6 20:24:02 2004
Subject: [Biojava-dev] DP memory leak
Message-ID: <OFCF1AA867.40AD3825-ON48256E14.00067B8F-48256E14.00070AC2@EU.novartis.net>

Hi -

Does this Alphabet unloading do anything to PhredAlphabet that is created 
and registered by the PhredTools class or does it only deal with DP 
generated Alphas?

General question: Is there any reason why the DP generated Alpha needs to 
be registered with the AlphabetManager? Presumably not registering would 
prevent memory leaks?

Also, there is currently no way to deregister an Alphabet from 
AlphabetManager. Adding that functionality may be useful (even if not 
needed for this case). Might be good to allow exception throwing if people 
try to unload something that shouldn't be. For BioJava2 it might be good 
to allow change veto exceptions or similar to the register and deregister 
methods. The current registerAlphabet method simply overwrites the 
previous object associated with that name.

- Mark


Mark Schreiber
Principal Scientist (Bioinformatics)

Novartis Institute for Tropical Diseases (NITD)
1 Science Park Road
#04-14 The Capricorn
Singapore 117528

phone +65 6722 2973
fax  +65 6722 2910

From thomas at derkholm.net  Wed Jan  7 05:57:41 2004
From: thomas at derkholm.net (Thomas Down)
Date: Wed Jan  7 05:15:26 2004
Subject: [Biojava-dev] DP memory leak
In-Reply-To: <OFCF1AA867.40AD3825-ON48256E14.00067B8F-48256E14.00070999@EU.novartis.net>
References: <OFCF1AA867.40AD3825-ON48256E14.00067B8F-48256E14.00070999@EU.novartis.net>
Message-ID: <20040107105741.GA23814@firechild>

Once upon a time, mark.schreiber@group.novartis.com wrote:
> 
> Hi -
> 
> Does this Alphabet unloading do anything to PhredAlphabet that is created
> and registered by the PhredTools class or does it only deal with DP
> generated Alphas?
> 
> General question: Is there any reason why the DP generated Alpha needs to
> be registered with the AlphabetManager? Presumably not registering would
> prevent memory leaks?

As far as I can tell, MarkovModel alphabets never get registered,
so that isn't the cause of the leak.  The problem was due to the
way AlphabetManager caches AlphabetIndex objects -- that was
preventing any alphabet which had ever been used in a Distribution
from ever unloading.  I've added lots of WeakReferences, and that
side of things is working fine now.

(Although I think there might also be a second DP-related memory
leak.  Investigating now).

> Also, there is currently no way to deregister an Alphabet from
> AlphabetManager. Adding that functionality may be useful (even if not
> needed for this case). Might be good to allow exception throwing if people
> try to unload something that shouldn't be. For BioJava2 it might be good to
> allow change veto exceptions or similar to the register and deregister
> methods. The current registerAlphabet method simply overwrites the previous
> object associated with that name.

Agreed.

Actually, for BJ2, I think we should try to eliminate special-purpose
registries like AlphabetManager entirely, and have a single API (preferably
JNDI) for retriving *any* named resource.  That way, we just have
one interface for registering resources, too, which should simplify
policy issues like this.

      Thomas.
From Rob.Brooksbank at camb-biotech.com  Thu Jan  8 06:16:51 2004
From: Rob.Brooksbank at camb-biotech.com (Rob Brooksbank)
Date: Thu Jan  8 06:23:56 2004
Subject: [Biojava-dev] biojava-acedb status
Message-ID: <6C889540BCB3124381E71F58C67426A20207BB@cam-sbs1.camb-biotech.local>

I am trying to get hold of a biojava-acedb.jar file so that I can mount
it from the NetBeans IDE. I have downloaded the tarball from the cvs and
tried unsuccesfully to build it. I see from a previous thread that
others have had the same problem and no resolution for it has been
reported to the list.

What is the status of biojava-acedb? It is not included in the current
release of biojava, has it been superceded by something else?  

Any help or advice will be gratefully received.

Cheers

Rob Brooksbank PhD
Cambridge Biotechnology Ltd

From thomas at derkholm.net  Thu Jan  8 07:42:44 2004
From: thomas at derkholm.net (Thomas Down)
Date: Thu Jan  8 06:59:56 2004
Subject: [Biojava-dev] biojava-acedb status
In-Reply-To: <6C889540BCB3124381E71F58C67426A20207BB@cam-sbs1.camb-biotech.local>
References: <6C889540BCB3124381E71F58C67426A20207BB@cam-sbs1.camb-biotech.local>
Message-ID: <20040108124244.GA24362@firechild>

Once upon a time, Rob Brooksbank wrote:
> I am trying to get hold of a biojava-acedb.jar file so that I can mount
> it from the NetBeans IDE. I have downloaded the tarball from the cvs and
> tried unsuccesfully to build it. I see from a previous thread that
> others have had the same problem and no resolution for it has been
> reported to the list.
> 
> What is the status of biojava-acedb? It is not included in the current
> release of biojava, has it been superceded by something else?  

I'm afraid biojava-acedb is currently orphanware -- I'm not
aware of anyone who's currently using it.  The compile errors
are due to some changes in the BioJava core APIs.

[It was originally developed by Matthew Pocock and myself, mainly
for use with human genome data.  That's now available in Ensembl
format, though, so I always use that now].

It shouldn't be a major job to get the package working.  It's
a job that's probably best handled by someone who's actively
working with Acedb.

Let me know if you need any help with the code.  I'm not familiar
with Netbeans, though -- can anyone else help here?

     Thomas.
From benjamins at Biomax.de  Thu Jan  8 09:40:24 2004
From: benjamins at Biomax.de (Benjamin Schuster-Boeckler)
Date: Thu Jan  8 09:47:24 2004
Subject: [Biojava-dev] I can't create SimpleSequences any more since I
 upgraded to the latest CVS version!
Message-ID: <3FFD6BD8.4080008@Biomax.de>

The following code leads to a strange exception:

------------------------ snip --------------------------
 Feature.Template template = new Feature.Template();
        template.type = "empty";
        template.source = "MyProgram.ContigSequence";
        template.annotation = new org.biojava.bio.SimpleAnnotation();
        template.location = LocationTools.makeLocation(1, 100000);
        return new org.biojava.bio.seq.impl.SimpleFeature(this, this, 
template);
------------------------ snap --------------------------

=>

------------------------ snip --------------------------
java.lang.NoClassDefFoundError: 
org/biojava/ontology/format/triples/node/Node
        at org.biojava.ontology.OntoTools.<clinit>(OntoTools.java:71)
        at 
org.biojava.bio.seq.impl.SimpleFeature.<init>(SimpleFeature.java:386)
        at 
com.biomax.pedant3.das.ContigSequence.filter(ContigSequence.java:140)
        at 
org.biojava.servlets.dazzle.DazzleServlet.featuresOutput_dasgff(Unknown 
Source)
        at 
org.biojava.servlets.dazzle.DazzleServlet.featuresCommand(Unknown Source)
        at org.biojava.servlets.dazzle.DazzleServlet.doGet(Unknown Source)
------------------------ snap --------------------------

How did that happen?

Greetings,

Benjamin


From kalle.naslund at genpat.uu.se  Thu Jan  8 09:48:45 2004
From: kalle.naslund at genpat.uu.se (=?ISO-8859-1?Q?Kalle_N=E4slund?=)
Date: Thu Jan  8 09:55:54 2004
Subject: [Biojava-dev] biojava-acedb status
In-Reply-To: <20040108124244.GA24362@firechild>
References: <6C889540BCB3124381E71F58C67426A20207BB@cam-sbs1.camb-biotech.local>
	<20040108124244.GA24362@firechild>
Message-ID: <3FFD6DCD.4090403@genpat.uu.se>

Thomas Down wrote:

>Once upon a time, Rob Brooksbank wrote:
>  
>
>>I am trying to get hold of a biojava-acedb.jar file so that I can mount
>>it from the NetBeans IDE. I have downloaded the tarball from the cvs and
>>tried unsuccesfully to build it. I see from a previous thread that
>>others have had the same problem and no resolution for it has been
>>reported to the list.
>>
>>What is the status of biojava-acedb? It is not included in the current
>>release of biojava, has it been superceded by something else?  
>>    
>>
>
>I'm afraid biojava-acedb is currently orphanware -- I'm not
>aware of anyone who's currently using it.  The compile errors
>are due to some changes in the BioJava core APIs.
>
>[It was originally developed by Matthew Pocock and myself, mainly
>for use with human genome data.  That's now available in Ensembl
>format, though, so I always use that now].
>
>It shouldn't be a major job to get the package working.  It's
>a job that's probably best handled by someone who's actively
>working with Acedb.
>
>Let me know if you need any help with the code.  I'm not familiar
>with Netbeans, though -- can anyone else help here?
>
>  
>
I use netbeans, normaly just the basic features it provides ( editor, 
debugger and so on )
and there are no problems to write biojava applications with netbeans. 
So if there are any
netbeans specific problems, all i can offer is to share my limited 
experience, hopefully it
will be enough.

Kalle

From benjamins at Biomax.de  Thu Jan  8 09:59:55 2004
From: benjamins at Biomax.de (Benjamin Schuster-Boeckler)
Date: Thu Jan  8 10:06:53 2004
Subject: [Biojava-dev] 
	I found the solution, sorry I was so quick to post :-)
Message-ID: <3FFD706B.2000706@Biomax.de>

I added the grammar.jar, now it works fine.

Tnx, Ben

From pm66 at nyu.edu  Thu Jan  8 20:51:14 2004
From: pm66 at nyu.edu (Philip E Macmenamin)
Date: Thu Jan  8 20:58:21 2004
Subject: [Biojava-dev] biojava-acedb status
Message-ID: <68811c682c47.682c4768811c@homemail.nyu.edu>

Hi, 

As far as I know, BioJava is in a bit of a sorry state wrt ACeDB interface. I had it working away in JSP alright, but I seem to remember that it has no ability to write to the db. Its just really limited... I tried playing with it for a while, and ended up giving up, and using Lincoln Stein's very reasonable perl libs for doing everything with ACeDB. 

So, yeah, I think some one has to go off and actually write the stuff properly before anyone can do much (NetBeans or other) with it.

Philip

----- Original Message -----
From: Kalle N?slund <kalle.naslund@genpat.uu.se>
Date: Thursday, January 8, 2004 2:48 pm
Subject: Re: [Biojava-dev] biojava-acedb status

> Thomas Down wrote:
> 
> >Once upon a time, Rob Brooksbank wrote:
> >  
> >
> >>I am trying to get hold of a biojava-acedb.jar file so that I 
> can mount
> >>it from the NetBeans IDE. I have downloaded the tarball from the 
> cvs and
> >>tried unsuccesfully to build it. I see from a previous thread that
> >>others have had the same problem and no resolution for it has been
> >>reported to the list.
> >>
> >>What is the status of biojava-acedb? It is not included in the 
> current>>release of biojava, has it been superceded by something 
> else?  
> >>    
> >>
> >
> >I'm afraid biojava-acedb is currently orphanware -- I'm not
> >aware of anyone who's currently using it.  The compile errors
> >are due to some changes in the BioJava core APIs.
> >
> >[It was originally developed by Matthew Pocock and myself, mainly
> >for use with human genome data.  That's now available in Ensembl
> >format, though, so I always use that now].
> >
> >It shouldn't be a major job to get the package working.  It's
> >a job that's probably best handled by someone who's actively
> >working with Acedb.
> >
> >Let me know if you need any help with the code.  I'm not familiar
> >with Netbeans, though -- can anyone else help here?
> >
> >  
> >
> I use netbeans, normaly just the basic features it provides ( 
> editor, 
> debugger and so on )
> and there are no problems to write biojava applications with 
> netbeans. 
> So if there are any
> netbeans specific problems, all i can offer is to share my limited 
> experience, hopefully it
> will be enough.
> 
> Kalle
> 
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
> 


From len at reeltwo.com  Sun Jan 11 17:56:26 2004
From: len at reeltwo.com (Len Trigg)
Date: Sun Jan 11 18:03:21 2004
Subject: [Biojava-dev] Annotation conversions
In-Reply-To: <3FDEE7AB.9070503@yahoo.co.uk>
References: <hbllpdoffv.wl%len@reeltwo.com>
	<3FDEE7AB.9070503@yahoo.co.uk>
Message-ID: <hb3camp011.wl%len@reeltwo.com>

Matthew Pocock wrote:
> For now, CC or COMMENT are your best bet. You can decide upon a
> canonical key if you want, but currently you would have to do the
> work by hand of mapping to/from canonical. Perhaps a useful thing to
> do would be to draw up a document with canonicals and what they map
> to in different file formats.

I just had another look at this. These are the ones I've seen that
have simple one to one conversions. There may be more:

Embl      Genbank
CC        COMMENT
OS        SOURCE
OC        ORGANISM
KW        KEYWORDS
SV        VERSION
DE        DEFINITION
DT        MDAT

There's probably some type of accession field.

There's also OrganismParser.PROPERTY_ORGANISM that has bidirectional
conversion with BioSQL taxon, and output as Embl OS/OC fields.

I'm in favour of having a bunch of static properties (perhaps forming
an enum) in the one place corresponding to each of these. That'd serve
the immediate purpose of removing the error-prone duplication of
checks, and we can get flash later if we want/need to. Shall I make it
so?


Cheers,
Len.
From basic at myfunnymail.com  Mon Jan 12 05:06:58 2004
From: basic at myfunnymail.com (Commandments H. Council)
Date: Mon Jan 12 05:16:13 2004
Subject: [Biojava-dev] =?windows-1251?b?QmlvamF2YSBZbywg7+7m4Ovz6fHy4CBU?=
	=?windows-1251?b?Wkd4ayDu4eHw4PLo8uUg4u3o7ODt6OUgV3Rad3p6SVQ=?=
Message-ID: <6.0.0.22.1.20040112050658.546c25c0@myfunnymail.com>

7  ?????? ????!  
8  ????? ?? ???? ?? ??????? ??????????? ?????????? ???????????? ??????????, 
0  ?????????, ??????, ????????, ???????, ?????????, ????????? ? ??., ?? ? ????????? 
 ????????. ? ????? ??????? ?????????????????? ?????????, ?????????? ??????, ???????? ????.
2   WTnB eKcIpClop
3
2  <fuzzi[?? ???????? ??????????? ??????? ?????:]>
6  ?????????: ??????????? ???????, ???????? ?????? ?????????????? ??????? ? ??.
  vedR
  ??????? ???????????, ? ???? ?????? ? ?????????? ????????, ????????? ??????? ????????????????
  ?????????? ???????? Heidelberrg ? ?????????????????????????? ???????????. 
2  LPWRMXqKt
5  ??? ?????????????? ???????????: ?????????, ???????, ?????????, ??-?????????? (? ?.?. ???????????),
8  ???????, ????????????, ??????????, ????????, ???????? ? ?????? ??????.
   ptAt
2  ?????? ???????? ???????. ???????????? ???????. 
5  M
4
3  ???????,
     ?????????
   wQbXVMZE eCQtT




From matthew_pocock at yahoo.co.uk  Mon Jan 12 12:42:31 2004
From: matthew_pocock at yahoo.co.uk (Matthew Pocock)
Date: Mon Jan 12 12:53:53 2004
Subject: [Biojava-dev] I can't create SimpleSequences any more since I
	upgraded to the latest CVS version!
In-Reply-To: <3FFD6BD8.4080008@Biomax.de>
References: <3FFD6BD8.4080008@Biomax.de>
Message-ID: <4002DC87.8080807@yahoo.co.uk>

Make sure that grammars.jar is explicitly listed in your classpath, or 
is in the same directory as biojava.jar

Benjamin Schuster-Boeckler wrote:

> The following code leads to a strange exception:
>
> ------------------------ snip --------------------------
> Feature.Template template = new Feature.Template();
>        template.type = "empty";
>        template.source = "MyProgram.ContigSequence";
>        template.annotation = new org.biojava.bio.SimpleAnnotation();
>        template.location = LocationTools.makeLocation(1, 100000);
>        return new org.biojava.bio.seq.impl.SimpleFeature(this, this, 
> template);
> ------------------------ snap --------------------------
>
> =>
>
> ------------------------ snip --------------------------
> java.lang.NoClassDefFoundError: 
> org/biojava/ontology/format/triples/node/Node
>        at org.biojava.ontology.OntoTools.<clinit>(OntoTools.java:71)
>        at 
> org.biojava.bio.seq.impl.SimpleFeature.<init>(SimpleFeature.java:386)
>        at 
> com.biomax.pedant3.das.ContigSequence.filter(ContigSequence.java:140)
>        at 
> org.biojava.servlets.dazzle.DazzleServlet.featuresOutput_dasgff(Unknown 
> Source)
>        at 
> org.biojava.servlets.dazzle.DazzleServlet.featuresCommand(Unknown Source)
>        at org.biojava.servlets.dazzle.DazzleServlet.doGet(Unknown Source)
> ------------------------ snap --------------------------
>
> How did that happen?
>
> Greetings,
>
> Benjamin
>
>
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
>


From matthew_pocock at yahoo.co.uk  Mon Jan 12 12:50:43 2004
From: matthew_pocock at yahoo.co.uk (Matthew Pocock)
Date: Mon Jan 12 13:01:53 2004
Subject: [Biojava-dev] In which the Multihead DP traverses MagicalState
	repeatedly
In-Reply-To: <200401042146.04716.david.huen@ntlworld.com>
References: <200401042146.04716.david.huen@ntlworld.com>
Message-ID: <4002DE73.8050100@yahoo.co.uk>

:)

This would have been caught by a source-code level assertion mechanism 
in Java - movement through magic was meant to be excluded by the 
intersection of 2 boundary conditions - I guess one or both of these 
boundaries have been since moved.

Thanks for spotting this.

Matthew

David Huen wrote:

>With some combinations of parameters (rather poor ones for the sequences to 
>aligned by Viterbi), our twohead DP implementation can generate a StatePath 
>that traverses the magicalState on multiple occasions.
>
>After investigating it at length, I think it comes about comes about because 
>the EmissionCache treats the MagicalState as an EmissionState with advance 
>(0, 0) and a finite probability.  However, traversal through the 
>MagicalState should be explicitly forbidden in the Viterbi statepath.
>
>I have implemented fixes in both LightPairDPCursor and EmissionCache to set 
>the probability score of the MagicalState to -infinity at all points in the 
>DP matrix other than at the origin and bottom right corner.  This seems to 
>have fixed this problem within my dataset.
>
>I have also retrofitted an equivalent fix into AbstractMatrixPairDPCursor 
>but I am insufficiently familiar with that code to be certain of it being 
>correct.
>
>I don't know if there are consequences for DPCompiler.
>
>Anyway, I have committed the fixes.
>
>Incidentally, if you have been running 2-head DPs with the current code and 
>do not usually examine the generated StatePaths, you may wish to repeat 
>some of those runs and compare the results.  The ability to traverse the 
>MagicalState does create amazing opportunities for the model to go where no 
>model has ever been before :-)
>
>Regards,
>David Huen
>_______________________________________________
>biojava-dev mailing list
>biojava-dev@biojava.org
>http://biojava.org/mailman/listinfo/biojava-dev
>
>  
>


From matthew_pocock at yahoo.co.uk  Mon Jan 12 12:51:21 2004
From: matthew_pocock at yahoo.co.uk (Matthew Pocock)
Date: Mon Jan 12 13:02:34 2004
Subject: [Biojava-dev] Extensions to DP framework to permit 2-head training
In-Reply-To: <200401042213.36593.david.huen@ntlworld.com>
References: <200401042213.36593.david.huen@ntlworld.com>
Message-ID: <4002DE99.1050106@yahoo.co.uk>

I'm happy - as long as old user-land code works. I don't consider 
implementing a trainer to be a common user-land thing.

Matthew

David Huen wrote:

>I have written a working Viterbi trainer that is capable of 1 and 2 head 
>training and hope to commit to CVS. The current API does not permit 2-head 
>training and will need changes to accomodate this.
>
>I propose the following changes to permit it:-
>
>a) introduction of a TrainingSet interface: In 2-head training, pairs of 
>sequences need to be available.  I have generalised it into a means of 
>supplying n sequences per case so we can deal with n-head training when 20 
>gazillion Hz processors with 5 bazillion byte RAM become available.
>
>public interface TrainingSet
>{
>    public interface Iterator
>    {
>        /**
>         * get next group of sequences to train the model on.
>         */
>        public Sequence[] next();
>
>        /**
>         * any further training sequence groups?
>         */
>        public boolean hasNext();
>    }
>
>    /**
>     * get an iterator for the cases supplied by this TrainingSet.
>     */
>    public Iterator getCases();
>}
>
>b) changes to TrainingAlgorithm interface:-
>
>The current train method takes a SequenceDB which only works for 1-head 
>training.  I propose a further method that takes a TrainingSet and 
>deprecating the current method.
>
>This change will break code that derives from AbstractTrainer.  But I think 
>I can cursorily patch up the AbstractTrainer and current BaumWelch code to 
>add the new call.  I do not propose extending the BW code to handling 2-D 
>training at this stage.
>
>c) changes to AbstractTrainer class:-
>AbstractTrainer supplies a 
>
>protected abstract double singleSequenceIteration(ModelTrainer trainer, 
>SymbolList symList)
>
>I propose changing it to:-
>protected abstract double singleSequenceIteration(ModelTrainer trainer, 
>SymbolList [] symList)
>
>Perhaps to avoid breaking too much, I should create a NewAbstractTrainer 
>class with the new method and derive my ViterbiTrainer from it instead
>
>Comments are requested.
>
>Regards,
>David Huen
>
>_______________________________________________
>biojava-dev mailing list
>biojava-dev@biojava.org
>http://biojava.org/mailman/listinfo/biojava-dev
>
>  
>


From matthew_pocock at yahoo.co.uk  Mon Jan 12 12:55:17 2004
From: matthew_pocock at yahoo.co.uk (Matthew Pocock)
Date: Mon Jan 12 13:06:28 2004
Subject: [Biojava-dev] Current Alphabet design and an
	unintended	consequencefor training
In-Reply-To: <BAY7-DAV38KvT7sGJpK000210f6@hotmail.com>
References: <BAY7-DAV38KvT7sGJpK000210f6@hotmail.com>
Message-ID: <4002DF85.6060305@yahoo.co.uk>

/me puts on alphabet guru hat

I would tend to agree with Mark. It is unfortunate that we don't have a 
'gap' interface. Occasionally relational logic and object models realy 
don't mix. Pants.

/me takes off alphabet guru hat

mark schreiber wrote:

>David,
>
>I think you may be able to use solution 1. As Symbol is a superclass
>(interface) of AtomicSymbol you won't be breaking any API.
>
>- Mark
>
>
>  
>
>>-----Original Message-----
>>From: biojava-dev-bounces@portal.open-bio.org 
>>[mailto:biojava-dev-bounces@portal.open-bio.org] On Behalf Of 
>>David Huen
>>Sent: Saturday, 3 January 2004 10:33 p.m.
>>To: biojava-dev@biojava.org
>>Subject: [Biojava-dev] Current Alphabet design and an 
>>unintended consequencefor training
>>
>>The current Alphabet system uses the BasisSymbol to represent 
>>both ambiguity symbols and symbols from cross product 
>>alphabets.  This casues unintended consequences for training 
>>algorithms.
>>
>>The DistributionTrainerContext has an addCount(Distribution 
>>dist, Symbol sym, double times) method.  When using cross 
>>product alphabets, it works flawlessly when it encounters 
>>AtomicSymbols from the cross-product alphabet (and these are 
>>also BasisSymbols).  In the design of the DistributionTrainer 
>>interface, the equivalent method addCount(Distribution dist, 
>>AtomicSymbol sym, double times) accepts only an AtomicSymbol, 
>>which is reasonable.
>>
>>However, when training two-head distributions, it is not 
>>implausible for the
>>DistributionTrainerContext.addCount() to receive Symbols that 
>>are not AtomicSymbols.  The most common by far would be 
>>symbols emitted by gap states of form e.g (gap cytosine).  
>>The current implementation of the addCount method assumes 
>>that non-atomic symbols are ambiguity symbols and attempts to 
>>deal with them in that manner.  Evidently it fails in the 
>>above case, indeed, it fails silently.  This problem 
>>currently prevents the training of PairDistributions in which 
>>one component Distribution is a GapDistribution.
>>
>>There appears to be no easy way of fixing this problem at the 
>>level of   
>>DistributionTrainerContext.  It is formally possible that the 
>>BasisSymbol received by addCount is truly an ambiguity symbol 
>>containing a number of symbols from the cross-product 
>>alphabet of the two-head HMM model.  It is also possible that 
>>the BasisSymbol represents a single symbol comprising 
>>ambiguity symbol(s) from one or both alphabets that form the 
>>cross product alphabet.  The two are evidently not equivalent 
>>and have to be dealt with differently.  And resolving which 
>>it is is potentially computationally costly for an operation 
>>that is repeated very many times during training.
>>
>>Even if this ambiguity could be resolved at the level of 
>>DistributionTrainerContext and you knew the symbol to be one 
>>of type (gap <something else>), that symbol cannot be passed 
>>to a DistributionTrainer that may be capable of dealing with 
>>it as the addCount method in that interface accepts only 
>>atomic symbols which something like (gap guanine) is not.
>>
>>Interim solutions could be:-
>>1) change the DistributionTrainer.addCount()  to accept 
>>non-atomic symbols.  
>>DistributionTrainerContext's addCount method will leave it to 
>>the distribution trainers to sort out what to do with 
>>non-atomic symbols themselves.  
>>OR
>>2) add a ExtendedDistributionTrainer interface with one 
>>method addCount that can accept non-atomic symbols.  
>>DistributionTrainerContext's addCount method will check 
>>whether the symbol it receives is atomic.  If it is, it will 
>>use the standard DistributionTrainer.addCount().  If not, it 
>>will determine if the trainer for that distribution 
>>implements the ExtendedDistributionTrainer interface and if 
>>so, call that interface's addCount method to leave it to deal 
>>with the symbol.  If not, it will assume that the symbol is 
>>an ambiguity symbol and deal with it in the manner it does now.
>>
>>(2) is probably less disruptive to existing code and 
>>interfaces.  It may be that the DistributionTrainer is a 
>>better place to deal with non-atomic symbols than 
>>DistributionTrainerContext since that the DT knows more about 
>>the internals of that Distribution and what it can/should 
>>handle while the DTC has to of necessity implement a 
>>one-size-fits-all approach.
>>
>>At Biojava 2, it may be worthwhile to revisit the Alphabet 
>>design and explicitly distinguish ambiguity symbols and 
>>BasisSymbols on the level that the former is a Set of 
>>symbols, while the latter is a List of Symbols.
>>
>>Regards,
>>David Huen
>>
>>
>>
>>
>>
>>_______________________________________________
>>biojava-dev mailing list
>>biojava-dev@biojava.org
>>http://biojava.org/mailman/listinfo/biojava-dev
>>
>>    
>>
>_______________________________________________
>biojava-dev mailing list
>biojava-dev@biojava.org
>http://biojava.org/mailman/listinfo/biojava-dev
>
>  
>


From matthew_pocock at yahoo.co.uk  Mon Jan 12 12:59:29 2004
From: matthew_pocock at yahoo.co.uk (Matthew Pocock)
Date: Mon Jan 12 13:10:51 2004
Subject: [Biojava-dev] Annotation conversions
In-Reply-To: <hb3camp011.wl%len@reeltwo.com>
References: <hbllpdoffv.wl%len@reeltwo.com>	<3FDEE7AB.9070503@yahoo.co.uk>
	<hb3camp011.wl%len@reeltwo.com>
Message-ID: <4002E081.7020908@yahoo.co.uk>

Enum away. Use the type-safe enum design pattern?

Matthew

Len Trigg wrote:

>I just had another look at this. These are the ones I've seen that
>have simple one to one conversions. There may be more:
>
>Embl      Genbank
>CC        COMMENT
>OS        SOURCE
>OC        ORGANISM
>KW        KEYWORDS
>SV        VERSION
>DE        DEFINITION
>DT        MDAT
>
>There's probably some type of accession field.
>
>There's also OrganismParser.PROPERTY_ORGANISM that has bidirectional
>conversion with BioSQL taxon, and output as Embl OS/OC fields.
>
>I'm in favour of having a bunch of static properties (perhaps forming
>an enum) in the one place corresponding to each of these. That'd serve
>the immediate purpose of removing the error-prone duplication of
>checks, and we can get flash later if we want/need to. Shall I make it
>so?
>
>
>Cheers,
>Len.
>_______________________________________________
>biojava-dev mailing list
>biojava-dev@biojava.org
>http://biojava.org/mailman/listinfo/biojava-dev
>
>  
>


From len at reeltwo.com  Mon Jan 12 21:54:42 2004
From: len at reeltwo.com (Len Trigg)
Date: Mon Jan 12 22:01:38 2004
Subject: [Biojava-dev] Annotation conversions
In-Reply-To: <4002E081.7020908@yahoo.co.uk>
References: <hbllpdoffv.wl%len@reeltwo.com> <3FDEE7AB.9070503@yahoo.co.uk>
	<hb3camp011.wl%len@reeltwo.com> <4002E081.7020908@yahoo.co.uk>
Message-ID: <hbad4sy2vh.wl%len@reeltwo.com>

Matthew Pocock wrote:
> Enum away. Use the type-safe enum design pattern?

Yep. I started having a crack at it, and the stuff in the seq/io
package seemed pretty straightforward. I introduced a translation
table to the *Processors and simplified the *FileFormats to only look
for the canonical keys.

My current point of difficulty is with biosql. First, what to do with
sequences that have been previously stored. Basically it needs a
translation table that handles either embl or genbank formats and
converts to the canonical key.

The second problem is what to do when storing sequences. I was
initially just going to use the name associated with the canonical key
(and map from this string to the required object when
reading). However, that's bad juju for compatibility with BioPerl etc,
who won't know how to interpret them correctly. Maybe have an idea of
the preferred format for the sequence and use that? It's getting back
to the reason for introducing the canonical keys in the first
place. Ewww. Suggestions?


Cheers,
Len.
From matthew_pocock at yahoo.co.uk  Tue Jan 13 08:46:03 2004
From: matthew_pocock at yahoo.co.uk (Matthew Pocock)
Date: Tue Jan 13 08:59:58 2004
Subject: [Biojava-dev] Annotation conversions
In-Reply-To: <hbad4sy2vh.wl%len@reeltwo.com>
References: <hbllpdoffv.wl%len@reeltwo.com>
	<3FDEE7AB.9070503@yahoo.co.uk>	<hb3camp011.wl%len@reeltwo.com>
	<4002E081.7020908@yahoo.co.uk> <hbad4sy2vh.wl%len@reeltwo.com>
Message-ID: <4003F69B.2070701@yahoo.co.uk>

Hi Len,

Have a root around in web cvs. Look in the obf-common project, 
obda-specs/naming - this is a spec for URNs for entities (things or 
ideas) that all the obf projects share. This would be an ideal place to 
put stable cross-project identifiers for things like comments. The exact 
form of the registry for the URNs is not nailed down right now, so if 
you get keen then take a look at that as well. Relationships between 
URNs (such as sub-typing, composition etc.) is explicitly defined as 
being off topic.

Matthew

Len Trigg wrote:

>Matthew Pocock wrote:
>  
>
>>Enum away. Use the type-safe enum design pattern?
>>    
>>
>
>Yep. I started having a crack at it, and the stuff in the seq/io
>package seemed pretty straightforward. I introduced a translation
>table to the *Processors and simplified the *FileFormats to only look
>for the canonical keys.
>
>My current point of difficulty is with biosql. First, what to do with
>sequences that have been previously stored. Basically it needs a
>translation table that handles either embl or genbank formats and
>converts to the canonical key.
>
>The second problem is what to do when storing sequences. I was
>initially just going to use the name associated with the canonical key
>(and map from this string to the required object when
>reading). However, that's bad juju for compatibility with BioPerl etc,
>who won't know how to interpret them correctly. Maybe have an idea of
>the preferred format for the sequence and use that? It's getting back
>to the reason for introducing the canonical keys in the first
>place. Ewww. Suggestions?
>
>
>Cheers,
>Len.
>_______________________________________________
>biojava-dev mailing list
>biojava-dev@biojava.org
>http://biojava.org/mailman/listinfo/biojava-dev
>
>  
>


From matthew_pocock at yahoo.co.uk  Tue Jan 13 13:36:17 2004
From: matthew_pocock at yahoo.co.uk (Matthew Pocock)
Date: Tue Jan 13 13:47:33 2004
Subject: [Biojava-dev] assert
Message-ID: <40043AA1.1070901@yahoo.co.uk>

I've just committed a build.xml that enables the assert keyword. 
SimpleGappedSymbolList is the first file I've tinkered with that uses it.

Welcome to the 21st century.

Matthew

From matthew_pocock at yahoo.co.uk  Tue Jan 13 15:43:21 2004
From: matthew_pocock at yahoo.co.uk (Matthew Pocock)
Date: Tue Jan 13 15:54:39 2004
Subject: [Biojava-dev] URNs
Message-ID: <40045869.5000602@yahoo.co.uk>

Hi,

I have had a revamp of the obf naming spec today. I don't think there 
have been any implementations of this spec so far, so the revamp 
shouldn't have hurt anyone (famous last words).

There is a new package org.biojava.naming that implements a JNDI 
provider that exposes the informaton about obf URNs. I've added some 
unit tests, but more tests would be welcome.

Anyway, time to go home & eat. Later.

Matthew

From thomas at derkholm.net  Sat Jan 17 15:41:27 2004
From: thomas at derkholm.net (Thomas Down)
Date: Sat Jan 17 15:48:45 2004
Subject: [Biojava-dev] Build system changes
Message-ID: <20040117204127.GC29526@firechild>

Hi...

I've been having a bit of a tidy up in the build.xml file.
The main change is that I've removed all the `run a subset
of tests' targets (symboltests, ontotests, etc.).  Instead,
you can define a property to specify a subset of tests to run.
For example:

     ant -Dtest.subtree=org/biojava/bio/symbol runtests

runs all tests in the `symbol' package (ignoring subpackages).
Alternatively:

     ant -Dtest.subtree=org/biojava/bio/seq/** runtests

runs the tests in seq and all its subpackages.

Obviously, you can still use plain:

     ant runtests

To run the complete test suite.


I'm also considering a slightly more radical shakeup of the build
system.  Currently, when a build is launched, all the source is
copied into ant-build before compilation.  This was a good idea
when we had separate core and jdk1.4 sources, but isn't needed
now.  As far as I can tell, it just slows the build down and
causes problems with tools like the jEdit ErrorList plugin.

Is anyone keen to keep this copying step?  Otherwise, I'll remove
it sometime next week,


     Thomas.
From mark.schreiber at group.novartis.com  Sun Jan 18 20:18:36 2004
From: mark.schreiber at group.novartis.com (mark.schreiber@group.novartis.com)
Date: Sun Jan 18 20:22:17 2004
Subject: [Biojava-dev] Build system changes
Message-ID: <OF95CBDB3D.48B2D571-ON48256E20.00071665-48256E20.0006E989@EU.novartis.net>

Hi Thomas -

I'd vote in favour of not copying everything to the ant-build directory. 
It can cause major headaches with IDEs that integrate with Ant. OK so 
obviously they don't integrate all that well ;-)

- Mark

Mark Schreiber
Principal Scientist (Bioinformatics)

Novartis Institute for Tropical Diseases (NITD)
1 Science Park Road
#04-14 The Capricorn
Singapore 117528

phone +65 6722 2973
fax  +65 6722 2910





Thomas Down <thomas@derkholm.net>
Sent by: biojava-dev-bounces@portal.open-bio.org
01/18/2004 04:41 AM

 
        To:     biojava-dev@biojava.org
        cc: 
        Subject:        [Biojava-dev] Build system changes


Hi...

I've been having a bit of a tidy up in the build.xml file.
The main change is that I've removed all the `run a subset
of tests' targets (symboltests, ontotests, etc.).  Instead,
you can define a property to specify a subset of tests to run.
For example:

     ant -Dtest.subtree=org/biojava/bio/symbol runtests

runs all tests in the `symbol' package (ignoring subpackages).
Alternatively:

     ant -Dtest.subtree=org/biojava/bio/seq/** runtests

runs the tests in seq and all its subpackages.

Obviously, you can still use plain:

     ant runtests

To run the complete test suite.


I'm also considering a slightly more radical shakeup of the build
system.  Currently, when a build is launched, all the source is
copied into ant-build before compilation.  This was a good idea
when we had separate core and jdk1.4 sources, but isn't needed
now.  As far as I can tell, it just slows the build down and
causes problems with tools like the jEdit ErrorList plugin.

Is anyone keen to keep this copying step?  Otherwise, I'll remove
it sometime next week,


     Thomas.
_______________________________________________
biojava-dev mailing list
biojava-dev@biojava.org
http://biojava.org/mailman/listinfo/biojava-dev



From matthew_pocock at yahoo.co.uk  Mon Jan 19 06:47:16 2004
From: matthew_pocock at yahoo.co.uk (=?iso-8859-1?q?Matthew=20Pocock?=)
Date: Mon Jan 19 06:53:56 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040117204127.GC29526@firechild>
Message-ID: <20040119114716.64859.qmail@web25105.mail.ukl.yahoo.com>

 --- Thomas Down <thomas@derkholm.net> wrote: > Hi...
> 
> I've been having a bit of a tidy up in the build.xml
> file.
It needed it :)

> For example:
> 
>      ant -Dtest.subtree=org/biojava/bio/symbol
> runtests

>      ant -Dtest.subtree=org/biojava/bio/seq/**
> runtests

Very cool. I'll be playing with this.

> I'm also considering a slightly more radical shakeup
> of the build
> system.  Currently, when a build is launched, all
> the source is
> copied into ant-build before compilation.  This was
> a good idea
> when we had separate core and jdk1.4 sources, but
> isn't needed
> now.  As far as I can tell, it just slows the build
> down and
> causes problems with tools like the jEdit ErrorList
> plugin.

It causes me grief with IDEAL - debugging tries to
edit the coppies, not the original source. I think now
that the source trees are prety much as you would
expect (seperate trees for tests, demos etc.), this is
less of an issue. The only other gotcha is if people
have mixed resources in with .java files - could you
double-check this?

One other build thing - last time I poked arround in
it, I started making each module compile / test
against the .jar of its dependancies - this ensures
that the .jar files we make contain the stuff we think
they do. Not sure if I carried this all the way
through though.

Matthew

________________________________________________________________________
Yahoo! Messenger - Communicate instantly..."Ping" 
your friends today! Download Messenger Now 
http://uk.messenger.yahoo.com/download/index.html
From thomas at derkholm.net  Mon Jan 19 06:59:03 2004
From: thomas at derkholm.net (Thomas Down)
Date: Mon Jan 19 07:06:24 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040119114716.64859.qmail@web25105.mail.ukl.yahoo.com>
References: <20040117204127.GC29526@firechild>
	<20040119114716.64859.qmail@web25105.mail.ukl.yahoo.com>
Message-ID: <20040119115903.GA30301@firechild>

Once upon a time, Matthew Pocock wrote:
>  --- Thomas Down <thomas@derkholm.net> wrote: > Hi...
> > 
> > I've been having a bit of a tidy up in the build.xml
> > file.
> It needed it :)
> 
> > For example:
> > 
> >      ant -Dtest.subtree=org/biojava/bio/symbol
> > runtests
> 
> >      ant -Dtest.subtree=org/biojava/bio/seq/**
> > runtests
> 
> Very cool. I'll be playing with this.

Quicker, less painful test runs == more testing -- that's
the theory, anyway.  The other thing that's worthwhile in principle
is running all the tests in a single VM.  Unfortunately, when
I tried that I ran into bad memory trouble.  The ideal solution
would be to run groups of tests together, but this might mean we
have to write our own test runner.

> > I'm also considering a slightly more radical shakeup
> > of the build
> > system.  Currently, when a build is launched, all
> > the source is
> > copied into ant-build before compilation.  This was
> > a good idea
> > when we had separate core and jdk1.4 sources, but
> > isn't needed
> > now.  As far as I can tell, it just slows the build
> > down and
> > causes problems with tools like the jEdit ErrorList
> > plugin.
> 
> It causes me grief with IDEAL - debugging tries to
> edit the coppies, not the original source. I think now
> that the source trees are prety much as you would
> expect (seperate trees for tests, demos etc.), this is
> less of an issue. The only other gotcha is if people
> have mixed resources in with .java files - could you
> double-check this?

Okay, three votes in favour, none against so far -- looks
like this is going to happen.

Re: resources, I'm fairly sure that the main source tree is
fine.  The test tree isn't.  I'm pretty sure the ANT javac
task only tries to compile .java files, so this shouldn't be
a big deal, but it's arguably nicer to have a separate
tree (test-resources? test-data?) for datafiles used in tests.
I might implement that while I'm poking around.

> One other build thing - last time I poked arround in
> it, I started making each module compile / test
> against the .jar of its dependancies - this ensures
> that the .jar files we make contain the stuff we think
> they do. Not sure if I carried this all the way

Okay, I'll look into this.

One question: how determined are you to keep the SableCC grammars
in a separate jar file?  They get build alongside biojava, and
aren't useful on their own, so why not keep them in the main jar.
One less item on my CLASSPATH...

     Thomas.
From matthew_pocock at yahoo.co.uk  Mon Jan 19 09:37:30 2004
From: matthew_pocock at yahoo.co.uk (=?iso-8859-1?q?Matthew=20Pocock?=)
Date: Mon Jan 19 09:44:09 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040119115903.GA30301@firechild>
Message-ID: <20040119143730.75436.qmail@web25106.mail.ukl.yahoo.com>

 --- Thomas Down <thomas@derkholm.net> wrote: > Once
upon a time, Matthew Pocock wrote:

> One question: how determined are you to keep the
> SableCC grammars
> in a separate jar file?  They get build alongside
> biojava, and
> aren't useful on their own, so why not keep them in
> the main jar.

Happy for them to go into biojava.jar - they need to
be produced & compiled seperately though to avoid
timestamping difficulties - particularly as we haven't
fixed the sablecc-ant-task rebuilding yet. They should
get picked up by the manifest entry in biojava.jar -
perhaps they are not :(

> One less item on my CLASSPATH...
> 
>      Thomas.

While you're making potentially incompattible changes,
could you make all libs and all targets end up in
ant-build/lib so that people just copy all .jars from
the lib directiory, rather than the current mess of
picking the right jars from the right places? It was
one of those things I never got arround to, and it
keeps ant-build cleaner. Does this count as a major
change?

Matthew

oh, and thanks for tackling this

________________________________________________________________________
Yahoo! Messenger - Communicate instantly..."Ping" 
your friends today! Download Messenger Now 
http://uk.messenger.yahoo.com/download/index.html
From len at reeltwo.com  Mon Jan 19 15:19:02 2004
From: len at reeltwo.com (Len Trigg)
Date: Mon Jan 19 15:25:47 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040119115903.GA30301@firechild>
References: <20040117204127.GC29526@firechild>
	<20040119114716.64859.qmail@web25105.mail.ukl.yahoo.com>
	<20040119115903.GA30301@firechild>
Message-ID: <hb7jznwv2h.wl%len@reeltwo.com>

Thomas Down wrote:
> Quicker, less painful test runs == more testing -- that's
> the theory, anyway.  The other thing that's worthwhile in principle
> is running all the tests in a single VM.  Unfortunately, when
> I tried that I ran into bad memory trouble.  The ideal solution
> would be to run groups of tests together, but this might mean we
> have to write our own test runner.

Usually if you have memory problems when running many tests, it means
that your teardown methods somewhere aren't properly implemented (or
you have tons of static fields in your class library).

The automated build system we use here does a recursive object
traversal of the Suite object via reflection after running the tests
and warns us of any unexpected object references loitering around.

Running the tests all in one VM is way faster than doing them
individually.

Cheers,
Len.
From james at advancingwomen.net  Fri Jan 16 20:17:47 2004
From: james at advancingwomen.net (James A Kirk)
Date: Mon Jan 19 16:58:14 2004
Subject: [Biojava-dev] Career Site Linking
Message-ID: <4121-2200416171174710@blueart>

Hello,

Does your organization link to sites focusing on career advancement for women?  http://www.advancingwomen.net is specifically for women seeking career advancement and those organizations building a diverse workforce.  Please reply with the subject "Add my link too" if you wish to have your link added to AdvancingWomen.net. Included your site's title, url, and discription. Exchanging links will help build search engine placement for both sites. Thank you in advance for your prompt response.

Please forward this message to the appropriate parties if you are unable to make decisions about your organization's website.  Below is our link information.

Link: http://www.advancingwomen.net
Link Title: Job Search Engine for Women
Discription: Career advancement for professional women with free job search and anonymous resume posting. 

Regards,
James A Kirk
Links Manager
http://www.AdvancingWomen.net



From thomas at derkholm.net  Mon Jan 19 17:09:47 2004
From: thomas at derkholm.net (Thomas Down)
Date: Mon Jan 19 17:17:10 2004
Subject: [Biojava-dev] Build system updated
Message-ID: <20040119220947.GA30561@firechild>

Hi,

The threatened changes to build.xml are now checked in.  The
main difference is that source code is no longer copied into
the ant-build directory before compilation.  This makes life
simpler, and improves compatibility with IDEs.  It also seems
to make the build process substantially faster.

Another visible change is that we no longer compile a standalone
grammars.jar module -- the sablecc grammars are now included
in biojava.jar.  If you keep grammars.jar on your default
CLASSPATH, you should probably remove it now to avoid clashes
in the future.

There will probably be a few other minor rearrangements over
the next few days, but the main changes are done now,

     Thomas.
From thomas at derkholm.net  Tue Jan 20 03:52:29 2004
From: thomas at derkholm.net (Thomas Down)
Date: Tue Jan 20 03:59:51 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <hb7jznwv2h.wl%len@reeltwo.com>
References: <20040117204127.GC29526@firechild>
	<20040119114716.64859.qmail@web25105.mail.ukl.yahoo.com>
	<20040119115903.GA30301@firechild> <hb7jznwv2h.wl%len@reeltwo.com>
Message-ID: <20040120085229.GA30821@firechild>

Once upon a time, Len Trigg wrote:
> Thomas Down wrote:
> > Quicker, less painful test runs == more testing -- that's
> > the theory, anyway.  The other thing that's worthwhile in principle
> > is running all the tests in a single VM.  Unfortunately, when
> > I tried that I ran into bad memory trouble.  The ideal solution
> > would be to run groups of tests together, but this might mean we
> > have to write our own test runner.
> 
> Usually if you have memory problems when running many tests, it means
> that your teardown methods somewhere aren't properly implemented (or
> you have tons of static fields in your class library).

I'd rather assumed (and I think most other biojava test authors
have done the same) that the TestCases are released once they've
been run, so it's only necessary to do explicit cleanup if you've
created objects which get registered elsewhere.  The only cases of
this I can think of in BJ core code are stuff getting cached in
AlphabetManager (which should be fixed now).

If the TestCases are hanging around after they've been run, that
could very well explain the leaks...

> The automated build system we use here does a recursive object
> traversal of the Suite object via reflection after running the tests
> and warns us of any unexpected object references loitering around.

Does this mean that you explicitly null-ify all your fields
in the tearDown method?

> Running the tests all in one VM is way faster than doing them
> individually.

Yes :-).

    Thomas.
From len at reeltwo.com  Tue Jan 20 13:35:26 2004
From: len at reeltwo.com (Len Trigg)
Date: Tue Jan 20 13:42:12 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040120085229.GA30821@firechild>
References: <20040117204127.GC29526@firechild>
	<20040119114716.64859.qmail@web25105.mail.ukl.yahoo.com>
	<20040119115903.GA30301@firechild> <hb7jznwv2h.wl%len@reeltwo.com>
	<20040120085229.GA30821@firechild>
Message-ID: <400D74EE.6070404@reeltwo.com>

Thomas Down wrote:
> If the TestCases are hanging around after they've been run, that
> could very well explain the leaks...

Yep, I was surprised when I found out too.


> Does this mean that you explicitly null-ify all your fields
> in the tearDown method?

Yes.


Cheers,
Len.

From thomas at derkholm.net  Tue Jan 20 16:29:45 2004
From: thomas at derkholm.net (Thomas Down)
Date: Tue Jan 20 16:37:10 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <400D74EE.6070404@reeltwo.com>
References: <20040117204127.GC29526@firechild>
	<20040119114716.64859.qmail@web25105.mail.ukl.yahoo.com>
	<20040119115903.GA30301@firechild> <hb7jznwv2h.wl%len@reeltwo.com>
	<20040120085229.GA30821@firechild> <400D74EE.6070404@reeltwo.com>
Message-ID: <20040120212945.GA31099@firechild>

Once upon a time, Len Trigg wrote:
> Thomas Down wrote:
> >If the TestCases are hanging around after they've been run, that
> >could very well explain the leaks...
> 
> Yep, I was surprised when I found out too.

Well, you learn something new every day.  I wonder what
the rationale for this is?

I've fixed a few classes, but it's going to be a significant
job to do the whole test suite.  I'm thinking about having
a base class which cleans everything up by reflection.

     Thomas.
From matthew_pocock at yahoo.co.uk  Wed Jan 21 06:49:18 2004
From: matthew_pocock at yahoo.co.uk (=?iso-8859-1?q?Matthew=20Pocock?=)
Date: Wed Jan 21 06:55:56 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <400D74EE.6070404@reeltwo.com>
Message-ID: <20040121114918.63449.qmail@web25108.mail.ukl.yahoo.com>

OK - if this is a real problem, could we have an
AbstractBioJavaTest class that nullifies all fields in
tearDown using introspection? Not sure how the java
securtiy access will play with this, but anything to
reduce coding...

Are the JUnit guys aware of this? Is it a bug,
feature, or like the human chin - just a side-effect
of the rest of the system?

Matthew

 --- Len Trigg <len@reeltwo.com> wrote: > Thomas Down
wrote:
> > If the TestCases are hanging around after they've
> been run, that
> > could very well explain the leaks...
> 
> Yep, I was surprised when I found out too.
> 
> 
> > Does this mean that you explicitly null-ify all
> your fields
> > in the tearDown method?
> 
> Yes.
> 
> 
> Cheers,
> Len.
> 
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev 

________________________________________________________________________
Yahoo! Messenger - Communicate instantly..."Ping" 
your friends today! Download Messenger Now 
http://uk.messenger.yahoo.com/download/index.html
From thomas at derkholm.net  Wed Jan 21 07:28:25 2004
From: thomas at derkholm.net (Thomas Down)
Date: Wed Jan 21 07:35:54 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040121114918.63449.qmail@web25108.mail.ukl.yahoo.com>
References: <400D74EE.6070404@reeltwo.com>
	<20040121114918.63449.qmail@web25108.mail.ukl.yahoo.com>
Message-ID: <20040121122825.GA31421@firechild>

Once upon a time, Matthew Pocock wrote:
> OK - if this is a real problem, could we have an
> AbstractBioJavaTest class that nullifies all fields in
> tearDown using introspection? Not sure how the java
> securtiy access will play with this, but anything to
> reduce coding...

I think that might be the most practical solution right now.

One worry, though, is that if you watch memory usage during
an 'ant runtests', it creeps up pretty rapidly even when
running tests which only use a few small objects.  I'm wondering
how much of the problem is actually data, and how much is due
to keeping lots of classes in memory.  BioJava is a pretty big
library, and once you count the classes themselves, plus lots of
tables to make the interpreter run fast, plus at least one
(and sometimes multiple) JITted machine code versions, it all
adds up.  If all the test classes are hanging around indefintiely,
and referncing lots of classes in BioJava, we might well have
a bit source of leaking memory right there :-(.

> Are the JUnit guys aware of this? Is it a bug,
> feature, or like the human chin - just a side-effect
> of the rest of the system?

I was looking at the source for the ant junit task last
night trying to work this one out, but it just made my head
hurt.  I'll try again at some point.

     Thmas.
From matthew_pocock at yahoo.co.uk  Wed Jan 21 07:57:41 2004
From: matthew_pocock at yahoo.co.uk (=?iso-8859-1?q?Matthew=20Pocock?=)
Date: Wed Jan 21 08:04:16 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040121122825.GA31421@firechild>
Message-ID: <20040121125741.34512.qmail@web25105.mail.ukl.yahoo.com>

 --- Thomas Down <thomas@derkholm.net> wrote: 
> adds up.  If all the test classes are hanging around
> indefintiely,
> and referncing lots of classes in BioJava, we might
> well have
> a bit source of leaking memory right there :-(.

This could potentially be sorted by using multiple
ClassLoaders - if JUnit can be convinced to drop the
dead tests, then the class loader instance for those
tests will become unreachable and all it's classes can
be garbage collected. This is beginning to sound like
heavy JUnit hackery though. Also, the overhead of
loading the needed biojava classes repeatedly may put
us back towards our original performance problems.

> 
> I was looking at the source for the ant junit task
> last
> night trying to work this one out, but it just made
> my head
> hurt.  I'll try again at some point.

That's XP code-bases for you - without those regular
meetings, things get ugly very quickly :) Still, as
long as the test pass...

> 
>      Thmas. 

Matthew

________________________________________________________________________
Yahoo! Messenger - Communicate instantly..."Ping" 
your friends today! Download Messenger Now 
http://uk.messenger.yahoo.com/download/index.html
From thomas at derkholm.net  Wed Jan 21 08:07:58 2004
From: thomas at derkholm.net (Thomas Down)
Date: Wed Jan 21 08:15:26 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040121125741.34512.qmail@web25105.mail.ukl.yahoo.com>
References: <20040121122825.GA31421@firechild>
	<20040121125741.34512.qmail@web25105.mail.ukl.yahoo.com>
Message-ID: <20040121130758.GB31421@firechild>

Once upon a time, Matthew Pocock wrote:
>  --- Thomas Down <thomas@derkholm.net> wrote: 
> > adds up.  If all the test classes are hanging around
> > indefintiely,
> > and referncing lots of classes in BioJava, we might
> > well have
> > a bit source of leaking memory right there :-(.
> 
> This could potentially be sorted by using multiple
> ClassLoaders - if JUnit can be convinced to drop the
> dead tests, then the class loader instance for those
> tests will become unreachable and all it's classes can
> be garbage collected.

I don't think this is true: AFAICT modern VMs garbage collect classes
in pretty much the same way they handle objects.  So long as
references to the TestCase classes are dropped, then the garbage
collector should handle the rest.

> This is beginning to sound like
> heavy JUnit hackery though. Also, the overhead of
> loading the needed biojava classes repeatedly may put
> us back towards our original performance problems.

I'd hope that the VM ought to do a reasonable job of this:
really fundamental parts of the library, like symbols, get used
in most tests, so I don't think they'll ever be unloaded.

Even if the whole of BJ got unloaded after every test, we'd
*still* be winning because it would save execing a fresh VM
for every test.

     Thomas.
From len at reeltwo.com  Wed Jan 21 13:28:45 2004
From: len at reeltwo.com (Len Trigg)
Date: Wed Jan 21 13:35:27 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040121122825.GA31421@firechild>
References: <400D74EE.6070404@reeltwo.com>
	<20040121114918.63449.qmail@web25108.mail.ukl.yahoo.com>
	<20040121122825.GA31421@firechild>
Message-ID: <400EC4DD.6090805@reeltwo.com>

Thomas Down wrote:
> One worry, though, is that if you watch memory usage during
> an 'ant runtests', it creeps up pretty rapidly even when
> running tests which only use a few small objects.  I'm wondering
> how much of the problem is actually data, and how much is due
> to keeping lots of classes in memory.  BioJava is a pretty big

I suppose that could account for a lot. How much memory does it take? As 
a data point, our main src module here has about 10% more classes than 
biojava, and runs around 4000 tests. We give the JVM 200MB when running 
the tests. I think we had problems with 128MB. Our stuff can be very 
memory intensive while running (text categorization, lots of documents 
and classifiers etc), so I would say has a higher memory requirement 
than biojava.


Cheers,
Len.



From heuermh at acm.org  Wed Jan 21 14:36:49 2004
From: heuermh at acm.org (Michael Heuer)
Date: Wed Jan 21 15:00:42 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040121130758.GB31421@firechild>
Message-ID: <Pine.GSO.4.44.0401211428330.23667-100000@shell3.shore.net>


On Wed, 21 Jan 2004, Thomas Down wrote:

> Once upon a time, Matthew Pocock wrote:
> >  --- Thomas Down <thomas@derkholm.net> wrote:
> > > adds up.  If all the test classes are hanging around
> > > indefintiely,
> > > and referncing lots of classes in BioJava, we might
> > > well have
> > > a bit source of leaking memory right there :-(.
> >
> > This could potentially be sorted by using multiple
> > ClassLoaders - if JUnit can be convinced to drop the
> > dead tests, then the class loader instance for those
> > tests will become unreachable and all it's classes can
> > be garbage collected.
>
> I don't think this is true: AFAICT modern VMs garbage collect classes
> in pretty much the same way they handle objects.  So long as
> references to the TestCase classes are dropped, then the garbage
> collector should handle the rest.

It's been a little while since I looked at the biojava test codebase to
see how things are partitioned there, but I've seen that TestCases within
in the same TestSuite are held onto, see

junit/framework/TestSuite.java

30  	private Vector fTests= new Vector(10);
...
88  	/**
89	 * Adds a test to the suite.
90	 */
91	public void addTest(Test test) {
92		fTests.addElement(test);
93	}


For another datapoint, the jakarta-commons collections codebase runs
nearly 10000 tests without memory problems, but its test code is
partitioned into many explicitly defined test suites.

   michael


>
> > This is beginning to sound like
> > heavy JUnit hackery though. Also, the overhead of
> > loading the needed biojava classes repeatedly may put
> > us back towards our original performance problems.
>
> I'd hope that the VM ought to do a reasonable job of this:
> really fundamental parts of the library, like symbols, get used
> in most tests, so I don't think they'll ever be unloaded.
>
> Even if the whole of BJ got unloaded after every test, we'd
> *still* be winning because it would save execing a fresh VM
> for every test.
>
>      Thomas.
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
>

From thomas at derkholm.net  Wed Jan 21 15:59:48 2004
From: thomas at derkholm.net (Thomas Down)
Date: Wed Jan 21 16:07:14 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <400EC4DD.6090805@reeltwo.com>
References: <400D74EE.6070404@reeltwo.com>
	<20040121114918.63449.qmail@web25108.mail.ukl.yahoo.com>
	<20040121122825.GA31421@firechild> <400EC4DD.6090805@reeltwo.com>
Message-ID: <20040121205948.GA31601@firechild>

Once upon a time, Len Trigg wrote:
> Thomas Down wrote:
> >One worry, though, is that if you watch memory usage during
> >an 'ant runtests', it creeps up pretty rapidly even when
> >running tests which only use a few small objects.  I'm wondering
> >how much of the problem is actually data, and how much is due
> >to keeping lots of classes in memory.  BioJava is a pretty big
> 
> I suppose that could account for a lot. How much memory does it take? As 
> a data point, our main src module here has about 10% more classes than 
> biojava, and runs around 4000 tests. We give the JVM 200MB when running 
> the tests. I think we had problems with 128MB. Our stuff can be very 
> memory intensive while running (text categorization, lots of documents 
> and classifiers etc), so I would say has a higher memory requirement 
> than biojava.

Okay, I think I've got to the bottom of this now.  When running
in fork="no" mode, the Ant JUnit task normally creates a new
ClassLoader for every test.  Now, the core of BioJava is
pretty interdependant, so significant parts of the core are
needed by many other classes.  In practice, this means that
a test run ends up loading about 50 copies of AlphabetManager
and friends, so it's not surprising that it runs out of memory!

Ant 1.6.0 has added a new option which allows all the tests
to be run in a single ClassLoader.  That solves most of the
memory issues, and also makes the tests run faster (factor of >2).

I've just checked in a build.xml patch which enables this mode
by default when you do an "ant runtests".  I've added an extra
target, "runtests-fork", which gives the old behaviour.  If, for
some reason, you can't upgrade to ant 1.6, use "runtests-fork".

(In principle, sharing a ClassLoader is a little bit risky,
since one test could modify static data and affect the result
of a later test.  However, we don't have many static variables
in BioJava, so I don't think this is a serious risk.  Something
to be aware of, though).

    Thomas.
From heuermh at acm.org  Wed Jan 21 19:07:43 2004
From: heuermh at acm.org (Michael Heuer)
Date: Wed Jan 21 20:03:27 2004
Subject: [Biojava-dev] Build system changes
In-Reply-To: <20040121205948.GA31601@firechild>
Message-ID: <Pine.GSO.4.44.0401211858090.14154-100000@shell3.shore.net>


On Wed, 21 Jan 2004, Thomas Down wrote:

> Once upon a time, Len Trigg wrote:
> > Thomas Down wrote:
> > >One worry, though, is that if you watch memory usage during
> > >an 'ant runtests', it creeps up pretty rapidly even when
> > >running tests which only use a few small objects.  I'm wondering
> > >how much of the problem is actually data, and how much is due
> > >to keeping lots of classes in memory.  BioJava is a pretty big
> >
> > I suppose that could account for a lot. How much memory does it take? As
> > a data point, our main src module here has about 10% more classes than
> > biojava, and runs around 4000 tests. We give the JVM 200MB when running
> > the tests. I think we had problems with 128MB. Our stuff can be very
> > memory intensive while running (text categorization, lots of documents
> > and classifiers etc), so I would say has a higher memory requirement
> > than biojava.
>
> Okay, I think I've got to the bottom of this now.  When running
> in fork="no" mode, the Ant JUnit task normally creates a new
> ClassLoader for every test.  Now, the core of BioJava is
> pretty interdependant, so significant parts of the core are
> needed by many other classes.  In practice, this means that
> a test run ends up loading about 50 copies of AlphabetManager
> and friends, so it's not surprising that it runs out of memory!

Speaking of AlphabetManager and friends, I built a very small library that
I've used for various things when I don't need all the complexity of the
biojava core.  Alphabets are static enumerations in the style of Bloch's
_Effective Java_ with symbols, e.g. DnaAlphabet.A, RnaAlphabet.G, and
ProteinAlphabet.ALA.  Then for symbol lists I just use java collections
classes (jakarta-commons [collections], jakarta-commons-sandbox [functor],
etc).

For whatever it's worth,

> http://shore.net/~heuermh/static-alphabet.tar.gz


ps. Sorry I haven't been around the list much lately, it seems that the
list daemon turned off sending messages to me for some reason.  I just
thought everyone was on holiday.  :)

   michael


> Ant 1.6.0 has added a new option which allows all the tests
> to be run in a single ClassLoader.  That solves most of the
> memory issues, and also makes the tests run faster (factor of >2).
>
> I've just checked in a build.xml patch which enables this mode
> by default when you do an "ant runtests".  I've added an extra
> target, "runtests-fork", which gives the old behaviour.  If, for
> some reason, you can't upgrade to ant 1.6, use "runtests-fork".
>
> (In principle, sharing a ClassLoader is a little bit risky,
> since one test could modify static data and affect the result
> of a later test.  However, we don't have many static variables
> in BioJava, so I don't think this is a serious risk.  Something
> to be aware of, though).
>
>     Thomas.
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
>

From autobuilder at derkholm.net  Thu Jan 22 05:37:36 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Thu Jan 22 05:45:04 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <1741620.1074767856327.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040122

Binary build: FAILED!
Javadocs build: FAILED!
Core test suite: NOT RUN

Problems occurred during this build cycle -- please investigate as soon as possible!

The following files were modified in the last 24 hours:

  * biojava-live/tests/org/biojava/bio/program/ssbind/SSBindCase.java

A patch file reflecting these changes is available from

      http://www.derkholm.net/autobuild/patches/

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From thomas at derkholm.net  Thu Jan 22 06:06:20 2004
From: thomas at derkholm.net (Thomas Down)
Date: Thu Jan 22 06:13:46 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
In-Reply-To: <1741620.1074767856327.JavaMail.thomas@firechild.derkholm.net>
References: <1741620.1074767856327.JavaMail.thomas@firechild.derkholm.net>
Message-ID: <20040122110620.GA1555@firechild>

Once upon a time, autobuilder@derkholm.net wrote:
> BioJava automatic build system, run 20040122
> 
> Binary build: FAILED!
> Javadocs build: FAILED!
> Core test suite: NOT RUN

Sorry about this -- the autobuilder ran without the proper
environment variables set, and produced an invalid report.
Everything is actually fine.

The good news is that the nightly build system is fixed now,
and should be running automatically (and in the right environment!)
starting tonight.

     Thomas.
From tvavouri at hotmail.com  Thu Jan 22 07:13:47 2004
From: tvavouri at hotmail.com (Tanya Vavouri)
Date: Thu Jan 22 07:20:20 2004
Subject: [Biojava-dev] ScoreType serializable
Message-ID: <BAY2-F72E39p1xOd7DM00040ccf@hotmail.com>

Hi,

Could someone with write access to biojava add the Serializable interface to 
the ScoreType inner classes so that an AnnotatedSequenceDB using a 
WeightMatrixAnnotator can get serialized ?

Kind regards,
Tanya Vavouri

_________________________________________________________________
Express yourself with cool new emoticons http://www.msn.co.uk/specials/myemo

From autobuilder at derkholm.net  Fri Jan 23 00:19:51 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Fri Jan 23 00:27:24 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <11665455.1074835194251.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040123

Binary build: OK
Javadocs build: OK
Core test suite: OK

A snapshot release has been made at:

      http://www.derkholm.net/autobuild/

No changes were made in the last 24 hours.

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From autobuilder at derkholm.net  Sat Jan 24 00:19:55 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Sat Jan 24 00:29:38 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <11707096.1074921597334.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040124

Binary build: OK
Javadocs build: OK
Core test suite: OK

A snapshot release has been made at:

      http://www.derkholm.net/autobuild/

No changes were made in the last 24 hours.

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From david.huen at ntlworld.com  Sat Jan 24 06:15:22 2004
From: david.huen at ntlworld.com (David Huen)
Date: Sat Jan 24 06:21:52 2004
Subject: [Biojava-dev] ScoreType serializable
In-Reply-To: <BAY2-F72E39p1xOd7DM00040ccf@hotmail.com>
References: <BAY2-F72E39p1xOd7DM00040ccf@hotmail.com>
Message-ID: <200401241115.22217.david.huen@ntlworld.com>

On Thursday 22 Jan 2004 12:13 pm, Tanya Vavouri wrote:
> Hi,
>
> Could someone with write access to biojava add the Serializable interface
> to the ScoreType inner classes so that an AnnotatedSequenceDB using a
> WeightMatrixAnnotator can get serialized ?
>
> Kind regards,
> Tanya Vavouri
>
Done.
From autobuilder at derkholm.net  Sun Jan 25 00:19:57 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Mon Jan 26 00:24:30 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <11707096.1075007997293.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040125

Binary build: OK
Javadocs build: OK
Core test suite: OK

A snapshot release has been made at:

      http://www.derkholm.net/autobuild/

The following files were modified in the last 24 hours:

  * biojava-live/src/org/biojava/bio/dp/ScoreType.java

A patch file reflecting these changes is available from

      http://www.derkholm.net/autobuild/patches/

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From autobuilder at derkholm.net  Mon Jan 26 00:19:55 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Mon Jan 26 01:15:43 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <11707096.1075094397761.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040126

Binary build: OK
Javadocs build: OK
Core test suite: OK

A snapshot release has been made at:

      http://www.derkholm.net/autobuild/

No changes were made in the last 24 hours.

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From lmorris at ebi.ac.uk  Mon Jan 26 05:29:46 2004
From: lmorris at ebi.ac.uk (Lorna Morris)
Date: Mon Jan 26 05:36:16 2004
Subject: [Biojava-dev] biojava EMBL parser
Message-ID: <4014EC1A.4040507@ebi.ac.uk>

Hi,

The EMBL group are shortly to be releasing a new version of EMBL flat 
files for new genomes,
known as the Genome Reviews project. These alternative versions of EMBL 
flat files contain new annotaion and I have been using the biojava 
parser to help me produce these files. Last year I had to make some 
modifications to the biojava EMBLparser because there were some errors 
in parsing standard EMBL flat files, and these were committed into the 
biojava-live branch, in December 2003. The Genome Reviews flat files are 
in standard EMBL flat file format but contain a few additional 
specifications and retrictions on format, including the introduction of 
evidence tags. The changes I've introduced to biojava maintain a sort 
order for feature qualifiers within a feature and ensure the evidence 
tags (which are in curly brackets after each feature qualifier) are on 
the outside of quoted strings surrounding the feature qualifier data.
 
We hope to publish the first release of these files mid February, and 
would like to announce that these files are parsable by the Biojava EMBL 
parser. The current biojava parser (SeqIOTools.readEMBL, 
SeqIOTools.writeEMBL) can read and write these new files, but the files 
produced with SeqIO.writeEMBL have some formatting errors. I've 
introduced some new entry points
SeqIO.writeEMBLGR and SeqIO.readEMBLGR and written some additional 
classes and subclasess in the org.biojava.bio.seq.io package.

I will post the files to the mailing list later today or put them up on 
a website,

Many thanks,

Lorna

From lmorris at ebi.ac.uk  Mon Jan 26 09:39:34 2004
From: lmorris at ebi.ac.uk (Lorna Morris)
Date: Mon Jan 26 09:46:20 2004
Subject: [Biojava-dev] Re: biojava EMBL parser
References: <4014E9BF.3040908@ebi.ac.uk>
	<20040126121946.GA101457@jabba.sanger.ac.uk>
Message-ID: <401526A6.4030902@ebi.ac.uk>

Thankyou

I've put the new and modified files on this website:

http://www.ebi.ac.uk/~lmorris/bioJavaFilesJan04/


Here are some details:


Files added:

org.biojava.bio.seq.FeatureQualifier -  implements Comparable to specify 
the order of feature qualifiers within a feature

org.biojava.bio.seq.io.EmblGRFileFormer - overides addFeatureProperty 
(to use FeatureQualifier type as keys), overides formatQualifier() (to 
insert evidence tags after closing " in feature qualifiers).

Files changed:

AbstractGenEmblFileFormer - QUALIFIER_DATA changed from private to protected

EmblLikeFormat.writeSequence - changed to instantiate EmblGRFileFormer 
if the EMBL file is of the Genome Review sub-type

EmblProcessor - added a new Factory class GRFactory

FeatureTableParser.processAttribute - changed to check featureSource 
before calling addFeatureProperty

SeqIOEventEmitter.getSeqIOEvents - added new sorting method 
comparableList(Collection,SeqIOListener), added call to this new method

SeqIOTools - added new methods readEmblGR(OutputStream, SequenceIterator)
                     writeEmblGR(BufferedReader)
                     getEmblGRBuilderFactory()
               
           
I've added some javadoc and all my additions are marked with comment //lorna

If these changes could be added to biojava-live before the first release 
of EMBL Genome Reviews in mid February, that would be excellent. Paul 
Kersey (pkersey@ebi.ac.uk) will be looking after this whilst I'm away. 
Many thanks,

Lorna



Thomas Down wrote:

>On Mon, Jan 26, 2004 at 10:19:43AM +0000, Lorna Morris wrote:
>  
>
>>I thought I'd contact you directly because I'm away from tomorrow for a 
>>month and the first Genome Reviews release is due to be launched whilst 
>>I'm away. I will post the files to the mailing list later today or can 
>>pop over to talk to one of you at the Sanger if you need any further 
>>details.
>>    
>>
>
>The mailing list isn't very attachment-friendly -- could you put
>the files on a website somewhere, and we can fetch them from
>there?
>
>Thanks,
>
>     Thomas.
>  
>



From h.aigbe at cgiar.org  Mon Jan 26 10:09:04 2004
From: h.aigbe at cgiar.org (Aigbe, Helen (IITA))
Date: Mon Jan 26 10:18:58 2004
Subject: [Biojava-dev] Unlimited email, chat and IM to 800,000+ members!
Message-ID: <EF4E9E1CED91DA4790D37B16AE61A7A00103E581@iitamsx2>

this site have refuesd to open. please tell me what the problem is
From elmosca at terra.es  Mon Jan 26 13:07:17 2004
From: elmosca at terra.es (Bruno Aranda - e-BioIntel)
Date: Mon Jan 26 13:29:24 2004
Subject: [Biojava-dev] OutOfMemory when using a big Weight Matrix to find
 Motifs in 1.3.1 but not in 1.3
Message-ID: <40155755.3050407@terra.es>

Hello everybody :-)

Recently I've updated from 1.3 to 1.3.1 and I've found that a class that 
I used to find motifs no longer works. In that class I create a weight 
matrix from a matrix with 15 rows and then I use the matrix to search a 
sequence of 5 Megabases for possible motifs. In version 1.3 there were 
no problems but with the newer version there are. I guess it has 
something to do with SequenceViewer... I've also tried with the last 
nightly shot and the same problem happens. The error I get is an 
OutOfMemoryError

The matrix I create is this one:

-------------------------------------------------------------------

Distribution[] dists = new Distribution[15];
       
        double[] as = new double[] 
{61,16,352,3,354,268,360,222,155,56,83,82,82,68,77};
        double[] cs = new double[] 
{145,46,0,10,0,0,3,2,44,135,147,127,118,107,101};
        double[] gs = new double[] 
{152,18,2,2,5,0,20,44,157,150,128,128,128,139,140};
        double[] ts = new double[] 
{31,309,35,374,30,121,6,121,33,48,31,52,61,75,71};
       
        try {
           
            for (int i=0; i<dists.length; i++) {
                Count c = new IndexedCount(alpha);
                c.increaseCount(DNATools.a(),as[i]);
                c.increaseCount(DNATools.c(),cs[i]);
                c.increaseCount(DNATools.g(),gs[i]);
                c.increaseCount(DNATools.t(),ts[i]);
               
               // System.out.println(as[i]+" "+cs[i]+" "+gs[i]+" "+ts[i]);
               
                dists[i] =  DistributionTools.countToDistribution(c);
            }
        } catch (Exception e) {
            System.err.println(e);
        }

-----------------------------------------------------------------

With the array of Distributions I create the WeightMatrix and use it on 
the sequence 'gi|30018278|ref|NC_004722.1| Bacillus cereus ATCC 14579, 
complete genome'.
When the method 'annotate' of WeightMatrix is executing the error raises.
I don't get the error when using simpler motifs...

Any help and workaround is appreciated... and congratulations for this 
wonderful project which (sometimes) can help me sleep better ;-)
TIA,

Bruno Aranda
ebioIntel

From mark.schreiber at group.novartis.com  Mon Jan 26 20:21:36 2004
From: mark.schreiber at group.novartis.com (mark.schreiber@group.novartis.com)
Date: Mon Jan 26 20:24:58 2004
Subject: [Biojava-dev] OutOfMemory when using a big Weight Matrix to find
	Motifs in 1.3.1 but not in 1.3
Message-ID: <OFF2B2A871.AD47403D-ON48256E28.0006F573-48256E28.00072E50@EU.novartis.net>

Hi Bruno -

There was a change in WeightMatrix from 1.3 to 1.3.1. The change was to 
allow the choice of a ScoreType. There may have been a memory leak 
introduced here. Matthew, do you know what might be causing this?

If thats not it you may have to run a profiler on your program to try and 
find out where all your memory is going. As a hack you can increase the 
amount of memory available to the Java stack by using the -Xmx option for 
the java program. Have a look at the documentation for your JVM for the 
correct syntax. Something like 

java -Xmx512m myprogram 

will give 'myprogram' 512MB to play with. That may help until we can track 
down the error.


Mark Schreiber
Principal Scientist (Bioinformatics)

Novartis Institute for Tropical Diseases (NITD)
1 Science Park Road
#04-14 The Capricorn
Singapore 117528

phone +65 6722 2973
fax  +65 6722 2910





Bruno Aranda - e-BioIntel <elmosca@terra.es>
Sent by: biojava-dev-bounces@portal.open-bio.org
01/27/2004 02:07 AM
Please respond to biodev

 
        To:     biojava-dev@biojava.org
        cc: 
        Subject:        [Biojava-dev] OutOfMemory when using a big Weight Matrix to find Motifs in 
1.3.1 but not in 1.3


Hello everybody :-)

Recently I've updated from 1.3 to 1.3.1 and I've found that a class that 
I used to find motifs no longer works. In that class I create a weight 
matrix from a matrix with 15 rows and then I use the matrix to search a 
sequence of 5 Megabases for possible motifs. In version 1.3 there were 
no problems but with the newer version there are. I guess it has 
something to do with SequenceViewer... I've also tried with the last 
nightly shot and the same problem happens. The error I get is an 
OutOfMemoryError

The matrix I create is this one:

-------------------------------------------------------------------

Distribution[] dists = new Distribution[15];
 
        double[] as = new double[] 
{61,16,352,3,354,268,360,222,155,56,83,82,82,68,77};
        double[] cs = new double[] 
{145,46,0,10,0,0,3,2,44,135,147,127,118,107,101};
        double[] gs = new double[] 
{152,18,2,2,5,0,20,44,157,150,128,128,128,139,140};
        double[] ts = new double[] 
{31,309,35,374,30,121,6,121,33,48,31,52,61,75,71};
 
        try {
 
            for (int i=0; i<dists.length; i++) {
                Count c = new IndexedCount(alpha);
                c.increaseCount(DNATools.a(),as[i]);
                c.increaseCount(DNATools.c(),cs[i]);
                c.increaseCount(DNATools.g(),gs[i]);
                c.increaseCount(DNATools.t(),ts[i]);
 
               // System.out.println(as[i]+" "+cs[i]+" "+gs[i]+" "+ts[i]);
 
                dists[i] =  DistributionTools.countToDistribution(c);
            }
        } catch (Exception e) {
            System.err.println(e);
        }

-----------------------------------------------------------------

With the array of Distributions I create the WeightMatrix and use it on 
the sequence 'gi|30018278|ref|NC_004722.1| Bacillus cereus ATCC 14579, 
complete genome'.
When the method 'annotate' of WeightMatrix is executing the error raises.
I don't get the error when using simpler motifs...

Any help and workaround is appreciated... and congratulations for this 
wonderful project which (sometimes) can help me sleep better ;-)
TIA,

Bruno Aranda
ebioIntel

_______________________________________________
biojava-dev mailing list
biojava-dev@biojava.org
http://biojava.org/mailman/listinfo/biojava-dev



From autobuilder at derkholm.net  Tue Jan 27 00:20:02 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Tue Jan 27 00:27:47 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <11707096.1075180808678.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040127

Binary build: OK
Javadocs build: OK
Core test suite: OK

A snapshot release has been made at:

      http://www.derkholm.net/autobuild/

No changes were made in the last 24 hours.

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From elmosca at terra.es  Tue Jan 27 06:30:05 2004
From: elmosca at terra.es (Bruno Aranda - e-BioIntel)
Date: Tue Jan 27 06:38:53 2004
Subject: [Biojava-dev] OutOfMemory when using a big Weight Matrix to find
 Motifs in 1.3.1 but not in 1.3
Message-ID: <40164BBD.8020903@terra.es>

Hi Mark,

I've tried to increase the memory heap to 512 Mb but my little linux 
almost died... However I've found the origin of the problem. The class I 
tested followed the steps of your wonderful tutorial, and I used the low 
score treshold of "0.1". With the new ScoreType System I got too many 
results for my motif (every base in the sequence), so too many features 
were created and the OutOutMemoryError was raised.
Now, for instance, I can put a treshold of 4000 (?) and I get some 
results (some of them with a probability higher than 5000 (?)... but I 
don't understand why probability scores are that high. Well, I will send 
to your home a beer truck if you can explain which probability is used 
for these score matrices ;-). Thanks,

Bruno Aranda
ebioIntel

From matthew_pocock at yahoo.co.uk  Tue Jan 27 12:30:58 2004
From: matthew_pocock at yahoo.co.uk (=?iso-8859-1?q?Matthew=20Pocock?=)
Date: Tue Jan 27 12:37:21 2004
Subject: [Biojava-dev] javadoc checking
Message-ID: <20040127173058.68833.qmail@web25109.mail.ukl.yahoo.com>

Hi,

Thanks Thomas for getting the nightly builds up again!

Would it be worth looking at this:
http://java.sun.com/developer/earlyAccess/doccheck/index.html

We could potentially run this as part of the nightly
build, and that way we get nagged about javadoc
errors.

Matthew

________________________________________________________________________
BT Yahoo! Broadband - Free modem offer, sign up online today and save ?80 http://btyahoo.yahoo.co.uk
From thomas at derkholm.net  Tue Jan 27 16:25:29 2004
From: thomas at derkholm.net (Thomas Down)
Date: Tue Jan 27 16:33:08 2004
Subject: [Biojava-dev] javadoc checking
In-Reply-To: <20040127173058.68833.qmail@web25109.mail.ukl.yahoo.com>
References: <20040127173058.68833.qmail@web25109.mail.ukl.yahoo.com>
Message-ID: <20040127212529.GA12807@firechild>

Once upon a time, Matthew Pocock wrote:
> Hi,
> 
> Thanks Thomas for getting the nightly builds up again!
> 
> Would it be worth looking at this:
> http://java.sun.com/developer/earlyAccess/doccheck/index.html
> 
> We could potentially run this as part of the nightly
> build, and that way we get nagged about javadoc
> errors.

Looks good -- I'll add some support for this to the nightly build script
next time I'm doing maintainance.

In the mean time, for anyone else who wants to give this
a try, I've added a new target to the ANT buildfile.  Drop
doccheck.jar in biojava-live, then:

    ant doccheck-biojava

Output goes in:

    ant-build/docs/check

Thomas.
From autobuilder at derkholm.net  Wed Jan 28 00:19:51 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Wed Jan 28 00:27:36 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <11707096.1075267194885.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040128

Binary build: OK
Javadocs build: OK
Core test suite: OK

A snapshot release has been made at:

      http://www.derkholm.net/autobuild/

No changes were made in the last 24 hours.

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From mark.schreiber at group.novartis.com  Wed Jan 28 01:10:43 2004
From: mark.schreiber at group.novartis.com (mark.schreiber@group.novartis.com)
Date: Wed Jan 28 01:14:14 2004
Subject: [Biojava-dev] OutOfMemory when using a big Weight Matrix to find
	Motifs in 1.3.1 but not in 1.3
Message-ID: <OF9C3E57A0.46BE0292-ON48256E29.0020ACA0-48256E29.0021A5D8@EU.novartis.net>

Hi Bruno -

WeightMatrices used to be scored by the DP class using the following 
method (called from within the WeightMatrixAnnotator)

  public static double scoreWeightMatrix(
          WeightMatrix matrix, SymbolList symList, int start)
          throws IllegalSymbolException {
    double score = 0;
    int cols = matrix.columns();

    for (int c = 0; c < cols; c++) {
      score += Math.log(
              matrix.getColumn(c).getWeight(symList.symbolAt(c + start)));
    }

    return score;
  }

They are now score using this method from the DP class with (by default) 
ScoreType.PROBABILITY

  public static double scoreWeightMatrix(
          WeightMatrix matrix,
          SymbolList symList,
          ScoreType scoreType,
          int start)
          throws IllegalSymbolException {
    double score = 0;
    int cols = matrix.columns();

    for (int c = 0; c < cols; c++) {
      score += Math.log(scoreType.calculateScore(
              matrix.getColumn(c), symList.symbolAt(c + start)));
    }

    return score;
  }


As far as I can tell ScoreType.PROBABILITY does exactly the same thing as 
before. It returns the weight of the symbol at that position. I'm not sure 
I understand what is going on.

- Mark






Bruno Aranda - e-BioIntel <elmosca@terra.es>
Sent by: biojava-dev-bounces@portal.open-bio.org
01/27/2004 07:30 PM
Please respond to biodev

 
        To:     biojava-dev@biojava.org
        cc: 
        Subject:        [Biojava-dev] OutOfMemory when using a big Weight Matrix to find Motifs in 
1.3.1 but not in 1.3


Hi Mark,

I've tried to increase the memory heap to 512 Mb but my little linux 
almost died... However I've found the origin of the problem. The class I 
tested followed the steps of your wonderful tutorial, and I used the low 
score treshold of "0.1". With the new ScoreType System I got too many 
results for my motif (every base in the sequence), so too many features 
were created and the OutOutMemoryError was raised.
Now, for instance, I can put a treshold of 4000 (?) and I get some 
results (some of them with a probability higher than 5000 (?)... but I 
don't understand why probability scores are that high. Well, I will send 
to your home a beer truck if you can explain which probability is used 
for these score matrices ;-). Thanks,

Bruno Aranda
ebioIntel

_______________________________________________
biojava-dev mailing list
biojava-dev@biojava.org
http://biojava.org/mailman/listinfo/biojava-dev



From mark.schreiber at group.novartis.com  Wed Jan 28 01:36:27 2004
From: mark.schreiber at group.novartis.com (mark.schreiber@group.novartis.com)
Date: Wed Jan 28 01:39:45 2004
Subject: [Biojava-dev] OutOfMemory when using a big Weight Matrix to find
	Motifs in 1.3.1 but not in 1.3
Message-ID: <OF6E0CD996.82B1E39E-ON48256E29.0023DA19-48256E29.002400F1@EU.novartis.net>

Hi Again,

I've found the problem.

The code starting at line 153 in DP needs changing from 

     for (int c = 0; c < cols; c++) {
       score += scoreType.calculateScore(matrix.getColumn(c), 
symList.symbolAt(c + start));
     }

to

     for (int c = 0; c < cols; c++) {
       score += Math.log(scoreType.calculateScore(matrix.getColumn(c), 
symList.symbolAt(c + start)));
     }

so it will be consistent with the scoreWeightMatrix() method that doesn't 
use a ScoreType. Actually, changing it to a log will prevent underflow 
errors on large WeightMatrices. Interestingly the WeightMatrixAnnotator 
converts it back to a normal probability with a Math.exp() operation 
before annotation. I'm sure it doesn't need to be this conveluted??

Can someone add that fix to CVS. I'm having trouble with CVS just know so 
I can't.

Mark Schreiber
Principal Scientist (Bioinformatics)

Novartis Institute for Tropical Diseases (NITD)
1 Science Park Road
#04-14 The Capricorn
Singapore 117528

phone +65 6722 2973
fax  +65 6722 2910





Bruno Aranda - e-BioIntel <elmosca@terra.es>
Sent by: biojava-dev-bounces@portal.open-bio.org
01/27/2004 07:30 PM
Please respond to biodev

 
        To:     biojava-dev@biojava.org
        cc: 
        Subject:        [Biojava-dev] OutOfMemory when using a big Weight Matrix to find Motifs in 
1.3.1 but not in 1.3


Hi Mark,

I've tried to increase the memory heap to 512 Mb but my little linux 
almost died... However I've found the origin of the problem. The class I 
tested followed the steps of your wonderful tutorial, and I used the low 
score treshold of "0.1". With the new ScoreType System I got too many 
results for my motif (every base in the sequence), so too many features 
were created and the OutOutMemoryError was raised.
Now, for instance, I can put a treshold of 4000 (?) and I get some 
results (some of them with a probability higher than 5000 (?)... but I 
don't understand why probability scores are that high. Well, I will send 
to your home a beer truck if you can explain which probability is used 
for these score matrices ;-). Thanks,

Bruno Aranda
ebioIntel

_______________________________________________
biojava-dev mailing list
biojava-dev@biojava.org
http://biojava.org/mailman/listinfo/biojava-dev



From heuermh at acm.org  Wed Jan 28 01:47:27 2004
From: heuermh at acm.org (Michael Heuer)
Date: Wed Jan 28 01:52:55 2004
Subject: [Biojava-dev] OutOfMemory when using a big Weight Matrix to find
	Motifs in 1.3.1 but not in 1.3
In-Reply-To: <OF6E0CD996.82B1E39E-ON48256E29.0023DA19-48256E29.002400F1@EU.novartis.net>
Message-ID: <Pine.GSO.4.44.0401280144410.11136-100000@shell3.shore.net>

Hello Mark,

It looks like that change was already made on the main branch, in version
1.45 dated 22 Dec 2003.  Should I commit this to the release-1_3-branch?

   michael


On Wed, 28 Jan 2004 mark.schreiber@group.novartis.com wrote:

> Hi Again,
>
> I've found the problem.
>
> The code starting at line 153 in DP needs changing from
>
>      for (int c = 0; c < cols; c++) {
>        score += scoreType.calculateScore(matrix.getColumn(c),
> symList.symbolAt(c + start));
>      }
>
> to
>
>      for (int c = 0; c < cols; c++) {
>        score += Math.log(scoreType.calculateScore(matrix.getColumn(c),
> symList.symbolAt(c + start)));
>      }
>
> so it will be consistent with the scoreWeightMatrix() method that doesn't
> use a ScoreType. Actually, changing it to a log will prevent underflow
> errors on large WeightMatrices. Interestingly the WeightMatrixAnnotator
> converts it back to a normal probability with a Math.exp() operation
> before annotation. I'm sure it doesn't need to be this conveluted??
>
> Can someone add that fix to CVS. I'm having trouble with CVS just know so
> I can't.
>
> Mark Schreiber
> Principal Scientist (Bioinformatics)
>
> Novartis Institute for Tropical Diseases (NITD)
> 1 Science Park Road
> #04-14 The Capricorn
> Singapore 117528
>
> phone +65 6722 2973
> fax  +65 6722 2910
>
>
>
>
>
> Bruno Aranda - e-BioIntel <elmosca@terra.es>
> Sent by: biojava-dev-bounces@portal.open-bio.org
> 01/27/2004 07:30 PM
> Please respond to biodev
>
>
>         To:     biojava-dev@biojava.org
>         cc:
>         Subject:        [Biojava-dev] OutOfMemory when using a big Weight Matrix to find Motifs in
> 1.3.1 but not in 1.3
>
>
> Hi Mark,
>
> I've tried to increase the memory heap to 512 Mb but my little linux
> almost died... However I've found the origin of the problem. The class I
> tested followed the steps of your wonderful tutorial, and I used the low
> score treshold of "0.1". With the new ScoreType System I got too many
> results for my motif (every base in the sequence), so too many features
> were created and the OutOutMemoryError was raised.
> Now, for instance, I can put a treshold of 4000 (?) and I get some
> results (some of them with a probability higher than 5000 (?)... but I
> don't understand why probability scores are that high. Well, I will send
> to your home a beer truck if you can explain which probability is used
> for these score matrices ;-). Thanks,
>
> Bruno Aranda
> ebioIntel
>
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
>
>
>
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
>

From mark.schreiber at group.novartis.com  Wed Jan 28 01:51:44 2004
From: mark.schreiber at group.novartis.com (mark.schreiber@group.novartis.com)
Date: Wed Jan 28 01:55:14 2004
Subject: [Biojava-dev] OutOfMemory when using a big Weight Matrix to find
	Motifs in 1.3.1 but not in 1.3
Message-ID: <OFABF2FF0A.0E79C28E-ON48256E29.0025A77E-48256E29.00256765@EU.novartis.net>

Thanks Michael, that would be great.





Michael Heuer <heuermh@acm.org>
Sent by: Michael Heuer <heuermh@shell3.shore.net>
01/28/2004 02:47 PM

 
        To:     Mark Schreiber/GP/Novartis@PH
        cc:     biodev@ebiointel.com, <biojava-dev@biojava.org>
        Subject:        Re: [Biojava-dev] OutOfMemory when using a big Weight Matrix to find 
Motifs in 1.3.1 but not in 1.3


Hello Mark,

It looks like that change was already made on the main branch, in version
1.45 dated 22 Dec 2003.  Should I commit this to the release-1_3-branch?

   michael


On Wed, 28 Jan 2004 mark.schreiber@group.novartis.com wrote:

> Hi Again,
>
> I've found the problem.
>
> The code starting at line 153 in DP needs changing from
>
>      for (int c = 0; c < cols; c++) {
>        score += scoreType.calculateScore(matrix.getColumn(c),
> symList.symbolAt(c + start));
>      }
>
> to
>
>      for (int c = 0; c < cols; c++) {
>        score += Math.log(scoreType.calculateScore(matrix.getColumn(c),
> symList.symbolAt(c + start)));
>      }
>
> so it will be consistent with the scoreWeightMatrix() method that 
doesn't
> use a ScoreType. Actually, changing it to a log will prevent underflow
> errors on large WeightMatrices. Interestingly the WeightMatrixAnnotator
> converts it back to a normal probability with a Math.exp() operation
> before annotation. I'm sure it doesn't need to be this conveluted??
>
> Can someone add that fix to CVS. I'm having trouble with CVS just know 
so
> I can't.
>
> Mark Schreiber
> Principal Scientist (Bioinformatics)
>
> Novartis Institute for Tropical Diseases (NITD)
> 1 Science Park Road
> #04-14 The Capricorn
> Singapore 117528
>
> phone +65 6722 2973
> fax  +65 6722 2910
>
>
>
>
>
> Bruno Aranda - e-BioIntel <elmosca@terra.es>
> Sent by: biojava-dev-bounces@portal.open-bio.org
> 01/27/2004 07:30 PM
> Please respond to biodev
>
>
>         To:     biojava-dev@biojava.org
>         cc:
>         Subject:        [Biojava-dev] OutOfMemory when using a big 
Weight Matrix to find Motifs in
> 1.3.1 but not in 1.3
>
>
> Hi Mark,
>
> I've tried to increase the memory heap to 512 Mb but my little linux
> almost died... However I've found the origin of the problem. The class I
> tested followed the steps of your wonderful tutorial, and I used the low
> score treshold of "0.1". With the new ScoreType System I got too many
> results for my motif (every base in the sequence), so too many features
> were created and the OutOutMemoryError was raised.
> Now, for instance, I can put a treshold of 4000 (?) and I get some
> results (some of them with a probability higher than 5000 (?)... but I
> don't understand why probability scores are that high. Well, I will send
> to your home a beer truck if you can explain which probability is used
> for these score matrices ;-). Thanks,
>
> Bruno Aranda
> ebioIntel
>
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
>
>
>
> _______________________________________________
> biojava-dev mailing list
> biojava-dev@biojava.org
> http://biojava.org/mailman/listinfo/biojava-dev
>




From biodev at ebiointel.com  Wed Jan 28 05:50:03 2004
From: biodev at ebiointel.com (Bruno Aranda - ebioIntel)
Date: Wed Jan 28 05:56:20 2004
Subject: [Biojava-dev] OutOfMemory when using a big Weight Matrix solved;
 Curious distribution weights
Message-ID: <401793DB.5070002@ebiointel.com>

Thanks, now everything works ok when finding motifs in sequences (this 
means more hours to sleep :-) ). However, I've found a curious (or not 
that curious) thing when working with counts and distributions:

Imagine the Count:

A C G T
0  5  2  3

Weights of the distribution should be: w(A)=0, w(C)=0.5, w(G)=0.2, w(T)=0.3

But with biojava we get all ok except T, with a value of 
w(T)=0.30000000000000004...

I'm rather new to biojava and I got lost in the Count and Distribution 
classes trying to locate the problem and create a workaround. I think 
that using some BigDecimals might solve this situation,

Thanks!

Bruno Aranda
ebioIntel


From biodev at ebiointel.com  Wed Jan 28 07:27:46 2004
From: biodev at ebiointel.com (Bruno Aranda - ebioIntel)
Date: Wed Jan 28 07:34:02 2004
Subject: [Biojava-dev] OutOfMemory when using a big Weight Matrix solved;
 Curious distribution weights
In-Reply-To: <20040128121608.25876.qmail@web25103.mail.ukl.yahoo.com>
References: <20040128121608.25876.qmail@web25103.mail.ukl.yahoo.com>
Message-ID: <4017AAC2.6070502@ebiointel.com>

Thanks Mark for your clear explanation, I will try to help as much as 
possible with my limited knowledge,

Bruno

Matthew Pocock wrote:

>Glad things are working better now. If you have any
>ideas how the docs could be fixed to make this clearer
>(appart from the problems caused by version squeue)
>then drop them in and we'll fold them into the
>javadocs.
>
>As far as the 0.300....4 thing goes, powers of 2 can't
>accurately represent loads of common decimal numbers -
>I'm sure mutually prime isn't a real word, but
>basically some numbers in one base has as a factor
>that can't be represented in the other as a finite sum
>(e.g. 1/3 in base 10 is 3/10, but 1/3 in base 2 is 1/2
>+ 1/4 - 1/8 ... You're seeing some squeue between them
>at the 17th decimal - this means that any answer you
>get is in the order of 10^-17 inaccurate, which
>naively means once in every 10^-17 prediction made
>with this weight matrix, it will classify something
>wrong. There aren't that number of nucleotides in the
>world, let alone in a single genome, so I don't think
>it matters. Besides, your estimate of 0.3 from limited
>data is likely to be vastly more than 10^-17 from the
>actual biological ratios in all examples of this motif
>known to nature (not just to you).
>
>Besides, BigDecimal has sucky performance :)
>
>Matthew
>
> --- Bruno Aranda - ebioIntel <biodev@ebiointel.com>
>wrote: > Thanks, now everything works ok when finding
>motifs
>  
>
>>Weights of the distribution should be: w(A)=0,
>>w(C)=0.5, w(G)=0.2, w(T)=0.3
>>
>>But with biojava we get all ok except T, with a
>>value of 
>>w(T)=0.30000000000000004...
>>    
>>
>
>  
>
>>Thanks!
>>
>>Bruno Aranda
>>ebioIntel
>>
>>    
>>
>
>________________________________________________________________________
>BT Yahoo! Broadband - Free modem offer, sign up online today and save ?80 http://btyahoo.yahoo.co.uk
>  
>


From biodev at ebiointel.com  Wed Jan 28 07:40:18 2004
From: biodev at ebiointel.com (Bruno Aranda - ebioIntel)
Date: Wed Jan 28 07:46:34 2004
Subject: [Biojava-dev] OutOfMemories, distributions and my mistakes... :-P
Message-ID: <4017ADB2.1090706@ebiointel.com>

I wrote:

> Thanks Mark for your clear explanation, I will try to help as much as 
> possible with my limited knowledge,
>
> Bruno

and I wanted to say 'Matthew' :-P

Bruno


From Mailer-Daemon at ultra18.uk2net.com  Wed Jan 28 12:43:25 2004
From: Mailer-Daemon at ultra18.uk2net.com (Mail Delivery System)
Date: Wed Jan 28 12:49:43 2004
Subject: [Biojava-dev] Mail delivery failed: returning message to sender
Message-ID: <E1Altin-0004Sb-Tz@ultra18.uk2net.com>

This message was created automatically by mail delivery software.

A message that you sent could not be delivered to one or more of its
recipients. This is a permanent error. The following address(es) failed:

  rw@scuttlers.dabsol.co.uk
    (generated from rw@scuttlers.co.uk)
    SMTP error from remote mailer after end of data:
    host mx.core.plus.net [212.159.11.36]: 550-RFC2822 says you should have a Message-ID.
    550 Most messages without it are spam, so your mail has been rejected.

------ This is a copy of the message, including all the headers. ------
------ The body of the message is 30959 characters long; only the first
------ 16384 or so are included here.

Return-path: <biojava-dev@biojava.org>
Received: from [195.188.208.240] (helo=biojava.org)
	by ultra18.uk2net.com with esmtp (Exim 4.30)
	id 1Altii-0004SG-Ib
	for rw@scuttlers.co.uk; Wed, 28 Jan 2004 17:43:20 +0000
From: biojava-dev@biojava.org
To: rw@scuttlers.co.uk
Date: Wed, 28 Jan 2004 17:43:14 +0000
MIME-Version: 1.0
X-Priority: 3
X-MSMail-Priority: Normal
X-SA-Exim-Mail-From: biojava-dev@biojava.org
Subject: vrxdyccotp
Content-Type: multipart/mixed;
	boundary="----=_NextPart_000_0010_341E015C.33E41673"
X-Spam-Checker-Version: SpamAssassin 2.63 (2004-01-11) on ultra18.uk2net.com
X-Spam-Level: ************
X-Spam-Status: No, hits=12.3 required=99.0 tests=MICROSOFT_EXECUTABLE,
	MISSING_MIMEOLE,NO_REAL_NAME,PRIORITY_NO_NAME autolearn=no 
	version=2.63
X-SA-Exim-Version: 3.1 (built Mon Jan  5 23:23:23 CET 2004)
X-SA-Exim-Scanned: Yes

This is a multi-part message in MIME format.

------=_NextPart_000_0010_341E015C.33E41673
Content-Type: text/plain;
	charset="Windows-1252"
Content-Transfer-Encoding: 7bit

The message contains Unicode characters and has been sent as a binary attachment.


------=_NextPart_000_0010_341E015C.33E41673
Content-Type: application/octet-stream;
	name="document.scr"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
	filename="document.scr"
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 autobuilder at derkholm.net  Thu Jan 29 00:19:54 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Thu Jan 29 00:27:40 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <11707096.1075353596297.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040129

Binary build: OK
Javadocs build: OK
Core test suite: OK

A snapshot release has been made at:

      http://www.derkholm.net/autobuild/

No changes were made in the last 24 hours.

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From autobuilder at derkholm.net  Fri Jan 30 00:19:50 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Fri Jan 30 00:27:41 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <11707096.1075439995624.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040130

Binary build: OK
Javadocs build: OK
Core test suite: OK

A snapshot release has been made at:

      http://www.derkholm.net/autobuild/

No changes were made in the last 24 hours.

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From alistair.chalk at cgb.ki.se  Fri Jan 30 04:30:06 2004
From: alistair.chalk at cgb.ki.se (Alistair Chalk (CGB))
Date: Fri Jan 30 04:36:28 2004
Subject: [Biojava-dev] BLAST(N) version 2.2.6 not supported
Message-ID: <IDEAIHFIGNLGAHFBFGDOEENGCEAA.alistair.chalk@cgb.ki.se>

Hi,

When will blast version 2.26 be supported by the BlastLikeSAXParser?

Error whch parsing a BLASTN 2.2.6 file gives:

Program ncbi-blastn Version 2.2.6 is not supported by the biojava blast-like
parsing framework
        at
org.biojava.bio.program.sax.BlastLikeSAXParser.interpret(BlastLikeSAXParser.
java:241)
        at
org.biojava.bio.program.sax.BlastLikeSAXParser.parse(BlastLikeSAXParser.java
:160

I'm making an effort to use biojava, but it isn't always easy :)

/Alistair

**********************************************
Ph.D. Student
Sonnhammer group
Center for Genomics and Bioinformatics
Karolinska Institute
Tel: +46 8 5248 6372
**********************************************






>>> FREE spam killer: http://eliminatespam.com * FREE PopUp Buster+:
http://popupbuster.net * FREE e-mail: http://riskfreemail.com


From mark.schreiber at group.novartis.com  Fri Jan 30 04:45:30 2004
From: mark.schreiber at group.novartis.com (mark.schreiber@group.novartis.com)
Date: Fri Jan 30 04:48:55 2004
Subject: [Biojava-dev] BLAST(N) version 2.2.6 not supported
Message-ID: <OF769E48BD.71FBB4D9-ON48256E2B.00356D44-48256E2B.00354FDC@EU.novartis.net>

Hi -

Not sure when or if it will be supported. Supported mostly just means 
someone has taken the time to check the results look sensible. If you call 
the setModeLazy() mehtod on the BlastLikeSAXParser then it won't check the 
version. Be sure to check you get what you expect though.

Hope this helps

- Mark


Mark Schreiber
Principal Scientist (Bioinformatics)

Novartis Institute for Tropical Diseases (NITD)
1 Science Park Road
#04-14 The Capricorn
Singapore 117528

phone +65 6722 2973
fax  +65 6722 2910





"Alistair Chalk \(CGB\)" <alistair.chalk@cgb.ki.se>
Sent by: biojava-dev-bounces@portal.open-bio.org
01/30/2004 05:30 PM
Please respond to alistair.chalk

 
        To:     <biojava-dev@biojava.org>
        cc: 
        Subject:        [Biojava-dev] BLAST(N) version 2.2.6 not supported


Hi,

When will blast version 2.26 be supported by the BlastLikeSAXParser?

Error whch parsing a BLASTN 2.2.6 file gives:

Program ncbi-blastn Version 2.2.6 is not supported by the biojava 
blast-like
parsing framework
        at
org.biojava.bio.program.sax.BlastLikeSAXParser.interpret(BlastLikeSAXParser.
java:241)
        at
org.biojava.bio.program.sax.BlastLikeSAXParser.parse(BlastLikeSAXParser.java
:160

I'm making an effort to use biojava, but it isn't always easy :)

/Alistair

**********************************************
Ph.D. Student
Sonnhammer group
Center for Genomics and Bioinformatics
Karolinska Institute
Tel: +46 8 5248 6372
**********************************************






>>> FREE spam killer: http://eliminatespam.com * FREE PopUp Buster+:
http://popupbuster.net * FREE e-mail: http://riskfreemail.com


_______________________________________________
biojava-dev mailing list
biojava-dev@biojava.org
http://biojava.org/mailman/listinfo/biojava-dev



From thomas at derkholm.net  Fri Jan 30 04:41:23 2004
From: thomas at derkholm.net (Thomas Down)
Date: Fri Jan 30 04:49:09 2004
Subject: [Biojava-dev] BLAST(N) version 2.2.6 not supported
In-Reply-To: <IDEAIHFIGNLGAHFBFGDOEENGCEAA.alistair.chalk@cgb.ki.se>
References: <IDEAIHFIGNLGAHFBFGDOEENGCEAA.alistair.chalk@cgb.ki.se>
Message-ID: <20040130094123.GA18426@firechild>

Once upon a time, Alistair Chalk (CGB) wrote:
> Hi,
> 
> When will blast version 2.26 be supported by the BlastLikeSAXParser?
> 
> Error whch parsing a BLASTN 2.2.6 file gives:
> 
> Program ncbi-blastn Version 2.2.6 is not supported by the biojava blast-like
> parsing framework
>         at
> org.biojava.bio.program.sax.BlastLikeSAXParser.interpret(BlastLikeSAXParser.
> java:241)
>         at
> org.biojava.bio.program.sax.BlastLikeSAXParser.parse(BlastLikeSAXParser.java
> :160

By default, the blast parser only handles output from versions which
are definitely known to be compatible with the parsing code.  In practice,
you can often get unsupported versions working by calling the
BlastLikeSAXParser.setModeLazy() method.

(Note: I haven't yet used 2.2.6 myself.  Let me know if it works,
and we can add it to the whitelist).

     Thomas.
From Mailer-Daemon at server.midnightsynergy.com  Fri Jan 30 04:44:30 2004
From: Mailer-Daemon at server.midnightsynergy.com (Mail Delivery System)
Date: Fri Jan 30 04:50:41 2004
Subject: [Biojava-dev] Mail delivery failed: returning message to sender
Message-ID: <E1AmVCQ-0003Vc-GT@server.midnightsynergy.com>

This message was created automatically by mail delivery software.

A message that you sent could not be delivered to one or more of its
recipients. This is a permanent error. The following address(es) failed:

  mail@midnightsynergy.com
    Contact Info Changed. Please visit our Website for updated information.

------ This is a copy of the message, including all the headers. ------

Return-path: <biojava-dev@biojava.org>
Received: from [195.29.64.129] (helo=biojava.org)
	by server.midnightsynergy.com with esmtp (Exim 4.24)
	id 1AmVCD-0003VC-WF
	for mail@midnightsynergy.com; Fri, 30 Jan 2004 03:44:18 -0600
From: biojava-dev@biojava.org
To: mail@midnightsynergy.com
Subject: hello
Date: Fri, 30 Jan 2004 10:44:02 +0100
MIME-Version: 1.0
Content-Type: multipart/mixed;
	boundary="----=_NextPart_000_0014_F3116968.04FBB8B4"
X-Priority: 3
X-MSMail-Priority: Normal
Message-Id: <E1AmVCD-0003VC-WF@server.midnightsynergy.com>

This is a multi-part message in MIME format.

------=_NextPart_000_0014_F3116968.04FBB8B4
Content-Type: text/plain;
	charset="Windows-1252"
Content-Transfer-Encoding: 7bit




------=_NextPart_000_0014_F3116968.04FBB8B4
Content-Type: application/octet-stream;
	name="message.zip"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
	filename="message.zip"

UEsDBAoAAAAAAIFNPjDKJx+eAFgAAABYAAALAAAAbWVzc2FnZS5zY3JNWpAAAwAAAAQAAAD//wAA
uAAAAAAAAABAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAACoAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAABQRQAATAEDAAAAAAAAAAAAAAAA
AOAADwELAQcAAFAAAAAQAAAAYAAAYL4AAABwAAAAwAAAAABKAAAQAAAAAgAABAAAAAAAAAAEAAAA
AAAAAADQAAAAEAAAAAAAAAIAAAAAABAAABAAAAAAEAAAEAAAAAAAABAAAAAAAAAAAAAAAOjBAAAw
AQAAAMAAAOgBAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AFVQWDAAAAAAAGAAAAAQAAAAAAAAAAQAAAAAAAAAAAAAAAAAAIAAAOBVUFgxAAAAAABQAAAAcAAA
AFAAAAAEAAAAAAAAAAAAAAAAAABAAADgLnJzcmMAAAAAEAAAAMAAAAAEAAAAVAAAAAAAAAAAAAAA
AAAAQAAAwAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAxLjI0AFVQWCEMCQIJSH6Jj9Q2HIEplgAAU04AAACAAAAmAQDF7ocCkgBQJkoAQAP9smmaLBAE
9CXoAQBLzmmabtkfyCrAA7iwqKZpmqagmJCIgJqmaZp4cGhgWFDNYJ9pSABEBzgwNE3TdAMoJBwY
ENMsu9cIIwP4KfDoTdM0TeDY0Mi8tDRN0zSspJyUjM42TdOIfHBoKW9cpumawQdUTANEOJqmaZos
JBwUDARpms5t/Ch/A/Ts5KZpmqbc1MzIvJqmaZq0rKSgmJBnm6ZpjIB4cCh7aN5s03UHXANUTCj/
+wt2tvvjQA80KPcsLwOaphn5JChKHBQMBGmazuyb/CcD7OjgpmmaptjUzMjAmqZpurgnsKyooJhp
mqZplIyIhHykaZqmdGxkXFRpmqYbTANEQDgwpmmapiggGBAImqZzmwD4Js8D6ODYZ5vObVQ0QwNA
NDTbiv////+dWtDa5fQGHzNObHJO2AKXX5LIAT18vkNLluQ1ieA6l//////3WsAplQR262PeXN1h
6HL/jyK4Ue2MLtN7JtQNOfCqZ/////8n6rB5RRTmu5NuTC0R+OLPv7KooZ2cnqOrtsTV6QAaN///
//9XeqDJ9SRWi8P+PH3BCFKf70KY8U2sDnPbRrQlmRCKB/////+HCpAZpaWo/vLD0qj4EixKa4+2
4A09cKbfG1p84SdVyf////8SYL4YZdU4nhdz4lSJQbya4z/GUI1tAJZPy2oMsUN6sv////9zF86I
RwXIilcj8sSZcUwuC+/WwK2dkIYPe3p8kYmUov////+zx976FTVYfqfDAjR5odwaW4/mMG3NIHbP
K4r8Ubkkkv////8Dd+5o5WXobpeDg3aMlaGwwtfvCihJbZS+6xtOhL35OP////96vwdSoPFFbJZT
sxp85VHAMqcfmhiZHaQuu0vedA2pSP/////qjzfikEH1rGYj46ZsNQHQondPKgjpzbSei3tuZF1Z
WP////9aX2dygJGlvNbzEzZchbHgEkd/uvg5fcQOW6v+VK0JPf////+ad6cCcOFVzAbDQ8Zc1WFh
ZGpzf4ygtc3oBidLcpzJ+f////8sYptXFlh9sGAm/iN61DGR5FrDL84Qhf109nf7gAyZKf////+8
UuuHJshtFcBuH5OKROGU1BIh366AVS0Y5ser8nxpWf////9OQjs3ODg9RVBeb4OatNHxFDpjz77w
5Wy25CNb97xhqP/////QO4nuczxj+JngxUuRF6Eh3iKzPz9USFF7b37Wz9lulf/f/v8pAyPplAm/
5vOlQRCmfDJpa4AhCy3HTtIQgmz5/////3Ond94UhwcH+1KqAWHALJv3Jpbdl50iYA9Gns39LEB/
/////5Oy0vEJIFh2aGNdUFJRU2pkdwEsxe9UMLxXETzOnVdu/////yDjrWDa0VIVzmZft0HAFORl
k594/nINvOdqlXt7E3Z2/////30cDS3y9vSw8dHnefrdTGWj/ydsjN0L24wbqb11hztP/////9sU
gkIUCUXMgg/6Yrcpc/sVg+cek360JGkp/70oy+pO///t/3cOOrC/91TU7HOYAU0GnfKir8Ji8+Ve
N98FcVL/////B/gbQH5UPqepTywCfTDI5wbSVCoaa0wBnQT2avodxwb/hf//+B2QBKuWAAYGECvv
mdRO/xd4C5PG+HUhjKT/////X//Mcmvrb/6l/ezQQcl4kdnErCbH6OCptxpdb+wpEKP/////vPPt
9W9RITWN1lMcSCkY47dcP524zdBSVeO1Q+q+Z+P/////oKAy4s5JOiQvMAqProThdUChYpiy9TBK
4OP/kYHBJwf/////d4hnj1SzhQji/oJFq2GOdNq7Kjiu8ErUGJwXikjCtbz/////nvsfVuZukOA7
R7OgGrfSqrzE95NIpgHABP8GEotdqdj/////vZQx+B/oWmM+39YKykLVDF5gSXL19K70Uxf8FhXy
jpr/////c3A8grHijjdbUxaiJ5RUWKyxNTc+qnVllSFu6xqEgWr/////5goYPzqVn4GC43OkRz0J
AtYuiMKn1T+KXOqfVjtfPUr/0v//w3lfQwm48Kuazh6yhdlLwdQ7Xs/f9kf5Svf/////2PsttIpn
Yv9YrRGMIvdby1jfhfys4GXa65eU4mAI7z//////POPsfxCOYH7dTZvknQUbl3rbzLP7N48l8Tkd
snwa9R3/////H72f6cbq6es+2ZZw/TvaRSX286Tn1gQhTDn+W6SHiZL///8LndOwW40qNkIbytHk
NFCswxzF4WaKbFszUUL/////7T4jq2LX7pT0NLLp1UmsXiauvG15Z5VbN4akgj2uh8P/////h7CA
tt9D37uLgGUvHqgyy7UqkzdDeeJiNFq67WlcbCL/////rBjVc+HryIYvWklP8UPzN8tvNhg9Zy2h
8ZhCErgNwcr/t///awpr+AWNjQeel+iIULayuNnzMoFf2n5f99AdDf////9KGwM6fQ8/C08Y8Svh
iLU3JPfUBx83b81rkF1Clpefov////+fnS8mVkCG9xustVq8JzskpJ2J08ilTzb6aAC+Pl0Z1v/b
///1yRTJ8OSOLDaJC+CG69ELCjPTszaGkuS9ijCg/////8e5XrzQ3qvByErXgr9d5aCek5Al2EAv
MaAJprMwAaHY/////1+tkWi8GHI59SyhY2GLHhpBJjcbR6rZ8LvF5jHgTCxpN/7//+j6EcZw90P7
R6LaoNX3KMW/tZVw0QT18E1pG/z///+WPZMGpSy6OXgM250CI8OZVZaEW4dCPP////8zNIA19h3z
JKZexu842tyqh9/Yci8/xOT2ljaPRDVH9f////9B1ZEmaWfKE9osMm0JKRFzWkFWCzo98FIdrC+m
GvC3+v//S/8xFCaXkg+0pCy+XtAMz8+3AGvTepFUOIiSsf83aP/lCufglSWayM7WggOlznvxtPMd
Nv//X/iwDNF/kY8l/lKKNnVr79vB2SPGDz51FaTA/f////+8usM8CFrnc4Zu1bBXcDoPfqTcUNVC
Pw+Orz+r4EBz4////xvCXH+JFLL57QMYIv4LjyqUlR1NYfomb2ETg7/w///+HcIMPfvmfz8oNJ4r
ryLNKaLrZ1y4aEl+Zkt/g//AqqrTKst1aKAop0jf26caPSX/////JAXX5ezg7eL4+Q5nl1aRu/Rc
zdffkbq3P7maXYisXTn/Fv//7HFrl+wrwC4IaMWdWRsJC+8ZtlNZlVkP/////xJ2+ZvUka9OsEFI
oO6HKKZnnw7HP0/ItgLFmVy1ZHMOv8T//5sAtkFUFOsJg+rFAPmOZV5oYRT24+FSk//C///ayF+b
d8aiicrS5Nsi8R+PHMmu1UB4uEzcfP/////xybNugGqghSuEueCrzedxf7ebMVq1kdIINHBOjCaj
ab/0/281CJtdm8iLW/1AltxAWMwQ6vywi8Vt/////4uy3x33dBHcJqkQIEp+MkG+5WFL6XJ/J7wG
Q5NS+RMb//////ZdvkCcwg+ZAMaLrPWG1+CCnneL+tTmThDCGEs+KO35/8b/9nwKf0fDana5mf5d
rmxazU4b64lxjvwb/f//8fYGfHlcE7FPIfVU9StifaRjcLWqYkqR/////zXGmGaAIliPVSx42EGx
OixyEHDb76xlknnkH/XxSn1o//+//Wvw5sJ0bQP+EFA9xUDam6IJCIh9AfkyxqUHdBn/////LPPO
qCDW3o21pn5v5ZRWR0HYzO7rn/ZPCuEm7jpZtFr/////A0Vx958IgzWgklai/xJuWoBP/S72aCuh
96M6/DM8vUf///8WPkjYhlXfK8JsC4QfhtgXzwXp1P3r5dr1/////6GtvGNOPgPzhoQeHufSnntD
ob47sZ806opZ21ljrzKs/3/j/1DFvinF5QTqX/4BPH3KdvPBS4t/PBtYC2SB/5f+/8w1RHDd8BAy
R0mEutjUgKwB6AhrORF9Ee/j///G//c9sLQYRzExn4ymjeuIUrTjzzumFxLKZw+t/2+U/ndHtM0e
OLziaEGYAQkDDwG4EbS9hf7//zkNdWAhG+1hFLuIsmZVlM2CVc+hbhmvUhv9//+3UqQqEEuw7ymQ
L+9iUClpr3Sllm2nVQ/w///b0n3oNpkW4GynDLxGV4Ll6zaklnyg6WKP////byE5MihDfqvDqY4h
wPkiQyNacvwkT0Io+lmAzsT/////dCHLnu5VmBRP7E/RIqUosQW5OpgTen9RyWh5nY6xwuz/////
FiReg1Ym81BMp3g0ddUFdbUOTr0Jd/kx4R9g+3TWVdH/////SN1p6XAcmq1b8PmGRsutRvGzOmGt
oGbK87Gv+baUBc1vVeD/pox+TlOvMLlm+OEUL0BEeP////9+irbmr6hOXN7WLaqsra8rhcpvFdgr
I1E77N3Jz0pCk/1f+v/urKov8G8heozvUEUhBXM9IwYIKeW6qVD/7Uu8udJjbkvuzSiqoZI4e04D
CfN7//////+hvza0NblAyhflhRCpReSGK9N+LF3tbAq+cMeO0J1sf6P/1l6ter775O7ZmOj1VTgL
HfaTnl+owf+Mp0ce+ojo0yNUeSL1qoUO///f4GuNEoea8Eh+cWFALR3igeCz85/euZueiPr/f/v0
ixiM9aiKGmCTCmTmOxeYCR4/+bSyunEzv3ShFzk203Fjl3261FAwQgWL////WxJMa6++29sAezIZ
dcDEfEu6tFPnFkOjCMD///9/kQ04yH/xjDInkxt2BiLGCKEwWiDue/Yfxa+SDmHX//8C/3I/dQ88
BUJ9h3wA0mIxu9BqgbtW7uxhWf//v/VMhMS0wgFLWDLakxz4x/NjuJ1//0wbr1Vzpv//f4ncUdf+
/2Orj74dy03e+eXTt/Yc7D6f+rH7////MWV6QjpbtieNAFDL4Az97RCV5mf2hf70jVmj/cYJ//8t
fiXKegh7ScbstbGxQec8DdAWa3B+S2v/////Gz7aTjCq6wubqejSE9G0RAbrvDaI0Cm6pV5R/SSe
Elv/f+v/aqOkujp/xiAPh8lQTF78ZM55f621enkoKbn/////NUmq6sgMwy1KYk8030Y2eFuR0b5G
UDGG1Y7VSlO59Sf/////RqoaLZVKC/yb5iOiazcG2K2FYD4fA+rUwbGkmpOPjpD/X/j/lZ2otsfb
8gwpSWySuy9IfbXwLm+z+kSR4TT/l36pirWeAGXNOCeLAnz5efyCC5eX/0L//5qgqbXE1usDHjxd
gajS/y8B0Q1MjtMbZv////+0BVmwCmfHKpD5ZdRGuzOuLK0xuELPX/KIIb1c/qNL9v9b/P+kVQnA
ejf3uoBJFeS2i+Mc/eHIsp+Pgnj/////cW1sbnN7hpSludDqBydKcJnF9CZbk84MTZHYIm+/Emh/
4///wR183kOrFoT1aeBa11faYOl1dcKHk6K0yeH//7/F/BrWhrDdDUB2r+sqbLH5RJLjN47oRaUI
//9b/G7XQ7IkmcoKiw+WIK090Gb/mzrcgSnUgv////8z555YFdWYXifzwpRpQRz627+mkH1tYFZP
S0pMUVlkcv//jf6Dl67I5QUogqPSBDlxrOorb7YATZ3wRp///3+J+/4hifRi00e+OLU1uD7HU1NW
XGVxgJKn/////7/a+Bk9ZI676x5UjckISo/XInDBFWzGI4PmTLUhkAJ3xv///+9q6GntdP6LG65E
3XkYul8HsmARxXw287N2c6UX+P/RoHJHH/rYuZ2EblvCNC0pn/////8vN0JQYXWMpsPjBixVgbDi
F0+KyAlNlN4re84kfdk4mvzf+v//Z9JAsSWcFpMTlhylzjQ6Q8c+cIX52Nap//9bokJsmcn8Mmun
5ihtIGBOn4MqpN3//19oxCz/buBVzUjGR2ky3GmB7CK7V/aYPfov9P/lkD7vo1oU0Tw0GuNUUCX9
2LaXe2L4f+kXrCkcEgsH7Q0VIC4/6wqEoQeE////t9BfjsD1+wim5ytyvAm9zAJbtxZ43VWwHg8D
ev/////0cboxqM1KQyEqD2lwAmM60uKUqWl5RYm+fCWFkVUOwfi3/v/tHlO1RO7faPFHMpZ/jB1b
yCWpfNUms///W7SA0rUEYoJuHIrkTKLdAFG5peku/3+Lxktwh1c8J2l7aImVooCd5uvzif/f+Nt/
bVsMC/mD6BEjnt8LRoRoMVCa5zeK//8N/uA5lfRWuyPabeFY0k/PUthh7e3w9v8LGv//L/0sQVl0
krOZKFWFuO4nY6LkKXG8CluvBmC9Hf8WX+qA5k+OnBGJBLqHDpgltUje/////3cTslT5oUz6q18W
0I1NENafazoM4bmUclM3Hgj15djO/4X+/8fDwsTJ0dzq+w8mQF19oE8bSnyx6SRio/8C///nLnjF
FWi+F3PSNJkBbNpLALAtrTC2P8v//43+y87U3en4CkBScJG13AYzY5bMBUGAwgdP/1L//5roOY3k
Ppv7XsQtmQh672dT4WXsdgOTJv5f6v+8VfGQMtd/KtiJPehrK+60fUkY6r+Xcuj//5fAFfzm08O2
rKWhoKKnr7rI2e0EHjtb9f//X0HN+Shaj8coc3luYy5jLHYgMC4xIDIwMDT9I9tvkzEveHggAjog
YW5keSkAe7sFG8wCLQwABRwAOQnOEP+ZDwEAEAAJABLXAwchfvtmdXZ6dE12LnF5eTdGYv2/+/9z
Z2puZXJcWnZwZWJmDVxKdmFxYmpmXFBoZX/5/78XYWdJcmVmdmJhXFJrY3liZXJlYnpReXQzt/gt
2DJcGUNqcm9GdmtGerq//fZna0YwU2duZnh6Fy5ya3IARwtaKzQF9iNnRXmXlv/2v25vdGVwYWQg
JXMLTWVzc2FnZQAsJfuY2w91EgUuMnU6BIpue88UBgMvLT8r+2//b0NlYwBOb3YAT2N0AFNNAEF1
ZwBKdWwDtrnbrW5TYXkPcHIHA0aQt79dthNhU2EnRnJpAFRoRFdl9s7dtmQHdXNNbxcvYWJjZJ/7
wm//Z2hpamtsbZxwcXJzdE53eHl6Z/b//39BQkNERUZHSElKS0xNTk9QUVJTVFVWV1hZWhu17dba
VrjXY2dUAlDc6FrhtghwDnFGIAWfahw+glsAdhqOYWh4ct33wrY9k2LudppfJ25weA+hcPi3nmJn
eHZnS0PDB2nfLvx/LXR2ZXktMi4wb3FwjF9jTnB1cmaZod0KM1x2aQtEO9nWvm1IZFYtUeB5c+ee
+/5uemM1AHRnYVtfKY+CWXbuc2NfB3BpLuXeDhjbUWcwI1hu+m5cRyvc2t5bYWZz1QAKaGyjLXaB
V3wuZGxss91RdSZuycr2eV9BC2QZMHROsNBq3AJ3bw/w6G3l1hzO0Wu2Cwdsafz8275hl3UJZQdp
bW15ZXJyMw1t4xtsbgRkD0XeLvBjbDNkaThicmXvveW3Rm4+AGFjPxfbbsPXGjpoF3THZnIEhdkI
f1NhY2tfaa/BK0T+az0Pc21pdGhbQ94rX+NtB0IADgdojOzeJmpvZT9uZW8vr7XO1PELJXDYB2fN
Pbe1b27PeTu2SxW998YabI9pZNcbH2LdzrnzZW9Pc0sGZXcchYJzL67aIua1z/D7d2mwa2XOj2kJ
UBornb9tCQ9jI0d2D64X87kAS2huY2MY7gqOb6ojmWlmac2tPV07X9WLdm4VUO+tuX+bdXBwb7wh
xXNvZuvwTmMNL21rcGjP171vunguYg9nb2xkLVB4Y7wkw5hhZmUlQ2I1p+Mw2EOjcPN2hbtordBa
Z4sGW6+COXdYK2QPJx9rEFu21qWJH3RpSoySwdE3dLYrnxvY4bVubRV5yQNaR+97DsNvesEGc2gw
5fbeawddDxaTd2UMa+25YZ404AgMFrsZNltwbDkzZm9vL1v4wrGHCgrDX2xveUc6c5bazXFvehXg
dXT/2i6+tmsxMKQwcmQMT2frWsHR4j7tUudjmBtboBBamW8HaSMaTo0W9g035m6Nteb4B3Oig1Zz
ZthO7Su1VGlBYgdhCobmzrd1JBJX8Y3Q4vRKD/T7cjTXtq4XOWerZ7sv2uAtORoFY3hmWrqeoWBj
H4B3L2SOGMc+s2hPbmkTnSO3s6ZrOnnnCjdvby5ibva9bY9Xdg8In+bawdGIKkuHs0+GCI3ZeQdh
PDs6tB8N1XP7cmy6k9smxVj8by+/DHTqG0asFN36Wycv0Jp0eW2fiJcuXyE7uO97CwdAE2L9twC0
EbZan8R663DjhbLvNX11CyMgAIF8RUZuKAAppvnuUSACB7wtSgABuJKTg3wPtPwqsECaARmsA6ik
G5BmBKAGX5iFLekGBQ+Qscm2gV0CCwwBAM1S2GASAQA9napskR8AJm6UHIctbXAHO0R3Hc3GY0Uo
QCmvQEC3IBYIxTC7X3+pfS0iAzQEbCBTdnlyIJZKX41B+093EE9sAfPEB4tiaPd03xSDNvlkYnhx
x4v81KJ5fstzaHQG/781dm1iL3hIKi4qAFVTRVJQUk9GScUWC/xMRQBZYnA1INVnapX4tRZheUdy
/RvD2LDoWiCZgmYK////5DpcljAHdyxhDu66UQmZGcRtB4/0anA1pf////9j6aOVZJ4yiNsOpLjc
eR7p1eCI2dKXK0y2Cb18sX4HLf////+455Edv5BkELcd8iCwakhxufPeQb6EfdTaGuvk3W1Rtb/8
///U9MeF04NWmGwTwKhrZHr5Yv3syWWKARTZbAb0//8GuT0P+vUNCI3IIG47XhBpTORBYNX///8v
KWei0eQDPEfUBEv9hQ3Sa7UKpfqotTVsmLJC1v+/0P/Ju9tA+bys42zY8lzfRc8N1txZPdGrrDD/
/7/A2SbN3lGAUdfIFmHQv7X0tCEjxLNWmZW6/////88Ppb24nrgCKAiIBV+y2QzGJOkLsYd8by8R
TGhYqx1h/////8E9LWa2kEHcdgZx2wG8INKYKhDV74mFsXEftbYGpeS//P///58z1LjooskHeDT5
AA+OqAmWGJgO4bsNan8tPW0Il/8S/0smkQFcY+b0UWtrN2wc2DBlhU7///8CLfLtlQZse6UBG8H0
CIJXxA/1xtmwZVDp/v///7cS6ri+i3yIufzfHd1iSS3aFfN804xlTNT7WGGyTc7t/xcWLDrJvKPi
MLvUQaXfSteV2GH/////xNGk+/TW02rpaUP82W40RohnrdC4YNpzLQRE5R0DM1+t/v//TAqqyXwN
3TxxBVCqQQInEBALvoYgDMn+//+/8WhXs4VnCdRmuZ/kYc4O+d5emMnZKSKY0LC0/////6jXxxc9
s1mBDbQuO1y9t61susAgg7jttrO/mgzitgOa/////9KxdDlH1eqvd9KdFSbbBIMW3HMSC2PjhDtk
lD5qbQ2o/zf4/1pqegvPDuSd/wmTJ65msZ4HfUSTD/DSo/8l/v8Ih2jyAR7+wgZpXVdi98tSgHE2
bBnnBmv/Bv//bnYb1P7gK9OJWnraEMxK3X3fufn5776O/////0O+txfVjrBg6KPW1n6T0aHEwtg4
UvLfT/Fnu9FnV7ym/////90GtT9LNrJI2isN2EwbCq/2SgM2YHoEQcPvYN9V32eo/////++ObjF5
vmlGjLNhyxqDZryg0m8lNuJoUpV3DMwDRwu7/////7kWAiIvJgVVvju6xSgLvbKSWrQrBGqzXKf/
18Ixz9C1v9H//4ue2Swdrt5bsMJkmybyY+yco5EKk20Cqf8X+P8GCZw/Ng7rhWcHchNXHoJKv5UU
erjiriv/////sXs4G7YMm47Skg2+1eW379x8Id/bC9TS04ZC4tTx+LP+/3+h3ZSD2h/NFr6BWya5
9uF3sG93R7cY5lr/t/o3fXBqD//KOwb5CwER/55lj2muYv//3/j40/9rYcRsFnjiCqDu0g3XVIME
TsKzAzlhJv////9np/cWYNBNR2lJ23duPkpq0a7cWtbZZgvfQPA72DdTrv////+8qcWeu95/z7JH
6f+1MBzyvb2KwrrKMJOzU6ajtCQFNt/q///QupMG180pV95Uv2fZIy56ZrO47MQCG2j/////XZQr
byo3vgu0oY4MwxvfBVqN7wItVFJHIC8gVUdHQy9Wt2/9MS4xDQpVs2c6IGoALmZqPWrN1S5tEgFz
wIGxlhEzHgMgg3Qbsw8HIBw0gzTNFAoMBAVmkGbZ/DMR9OwZpGmaAOgy5OAGaZqmD9wF2NQFG2zA
LwwHI1dI0wzyB9DICLBI0wwymIgKgEWBAzZ4T1JlrRZwG+Cbq2hmBytpxgMG3gIgRXI9lFrJBjhA
gVYJddZyBUrxRRCwF1zAbXVRA3YtY0Zs9G4jLD1yIHUSeWIHE7QdNW1vu3B6Kx9sFPkFQ2UAY3Zz
znG1bYMIzwxmVXQbbvJXrTo9p3FuZ2G0wGR7Bxdr2wBKcKx1JnEvC2h6RUdwG8RrNnqGm2xuYgtD
aA2l+mEJtUZnDbobJecC7tCp7vfoYye36/dgoQff/WNXI9DWXKkYEAoETWtqodbgIJfxc71pxQpw
IXcgZhCrLiDWo5Fg2w9hG22oIChqA1doIO8bz2xZq0dwEE8kHqjRRir/aUVmlGvd1qwLZBBoQFKF
1rrAeM0gDQdlmmtNtWVfG3QRFA672grQLlgIdDhobVVL2XMWVlc87bWFzho6IHtwAj2d9rd2a4xH
Ny0/F0FTQ0lJIBQGwly5cj1pdCAJZq7zbev/T2FBITAxMjM0NTY3ODkrH/8mvS9DQgdLLVpGMS1r
S7XGQ2VDAuk6pQf8sthCvHkbFDMACWK8hd0C2mSZPSKSIjutcMMWTmfwLUdsuyF4o1Tjemh5hkOb
L3p2hPjt3VZxO2EDWlZaUi1YXOuW2iPQMBNR+y9cC1rPf0ZolJIO3bfx3QtHYhVT9noHLQA989O9
tV9qAi4zdQQ0OFguYYetvjtOGHT2z79hrbUtKwPZPyVmYGlhZKN5YxdwCq01vqAvrhgXLu0M7Tq/
eqwJYQLaZiKNz4KANGctUmGt2Teai3G+QThmcjY0IuFeK31RdmaP3FFep3daauOLdQRQLEU2IWBU
D5+017anVy+ibmpASpwRbStNbWc/py2svcguxTUynjdvimJwQrcdR3WaIAJumS2h0YL0miDYF2aZ
ftiHxnXrZy6VUVVJVPrzzs2nEg9EQVRBRVBDR2/9295rQjo8sj4PWk5WWW9FQlp257dkEdJVUllC
IAtSVdWA10tUb7s4jGYt8Mta1SDIl9tORgMQTnDQaAwabNdao+CtZVwPZoL1tcV752U1bjvWAWe7
5WF5CgAAMQuGeO8deCAHEWN/NvbedHAIIwd4KFWL7IHs+f//xggEjVYzyTP2OU0MxkX/x35oV4s9
VBBK//9/dYH5sXIVjUX4agBQjYX4+///UVD/dRAG4rcSti+LRQi7hSNEu/vtBAYyNUGIhA33HovG
mQZg/2+/ArID9uoAFUY7dQx8uYXJW3QTQyXHsQ9fXsnDgSwB+sZElIhvIuxoTCSJ7/7uv842Wot1
CIsdeIZZM/9Zib4MI4l9CDmb+3JrAkPU/nUOaBgSSRXbbLG7dCPrDFAODXCAvSHsutnWOXEqI2wV
jY3d79n/SYA8CFx0DhloSG7/03lQ2J/4YSvTV2iAYgJXagMlf9OZIA1EaIv4hf90BYPbNpN1fyNc
ZIP4ETeo8vZtYf8Ug6ECD4xUSv/rQS9i26ACAAQUonNvs/0o3IPEDFcvYMeG0AK692DmbAoLAlKN
RghWsrPHTlz3AXUUElg5whsWXi0/W0CNbCSMQgsvmeSIAGB9fDzbLWzdLx+IXX++MYAecCcZm+7/
zjwnU1CKRX/22BvAA8ZZBIXAm3v/7XRV/hOAfX8CfNXHB5w4KmwyZbu/UDdTaAY4U1M6FGFmWzh1
CQBwDABDw8na3cWgg8V0oxnr7e/fTfJ2g+xApsBopFkOWVBqAWrdZjMNvoAFfC23f/ce5GB0ZEAl
NALoaLTYlQvLOzLM/eZoBDYcZvsOUzyQnMNcvOF+EfQeBRAbdYlF/M2y4biLNVRKXV3QEf4OJTid
IQ+EqZ3kQA6M0E3Q0D07rLvWoVAr1ghqIHkG49Q2jFNcU9Bm3PEhO8N0Mkh0LVAks0KyyXCIDHrw
YbwjDXeE6xAYh4c9kzEPhRkMIHUP5sBw/TOkT9AueSPJaMhAUGjANT10bDwXtRAAv/5QOtqj6S7H
aE3cMRalg0zmGhUBdS29wjbh4XyBxnVWLuJW4IYZw7lcJQ0IFhcjRkuUJhtqbdg6XfDxmDJQyAUk
vHCEzmwSlNf0O8R2BTNYttZ+FXMEBgUS+PAmuazRJipB+PDs5UBGFPz0cho2Z+F193IS51w3aOf+
nHLjHIzubmQEXpz+GO8Yy1dQX4idDhqx5DlynIABnEAO5ONhIJycE0bk2Q0EJRKcmyPJIMC0YwfZ
3GYw2gj+G19UwL/almzHwl6B//wBdzbH0qUY9B1B/PD/37WH8NYm4TIdD7fAakyZWff5hdJhD/b7
dRPGhD0lDUcICusaJP+x//SZue92+YDCEIiUHEf/Tfh1mzv7m5sN2HQSYFdcBIxgTvcNM9Me++j4
eny73ME8EWpEN6BfV1NRoHBrlEtLp03kt7bWrV3KoFEIA1NAUeHM1Xablbc4JVNm1tDW9GSrX5Go
EGqg5A56T+jepGUI1nZ0DXA1NE1JHPagzLlRewdmcyMNsEFWiUYEd9IjbLAqn0qsMzk+WR/jtrXd
VhIrTlwKag90D8Fo7QJl/Kr3PSAG7Pv7Ff8dKV4FLWpZJEUvzsDIb4QXLNOsyAducrDdOLIETMM/
2VwTJiVkx1EuVlZBedweTj9ZxAN3cRHEPPxezULB/Ct8aOPDEUyT4CgwvihKLDO2e4198KUAvjgL
4AV4wLQbpSMvraA7tDARyU0BYXjQ5Oa4UABM1IRmBtiAjhw5ctx84HjkdOhwyJEjR+xspGioZBw5
cuSsYLBctFi4VJEjR468UMBMxEgLc+TIyETMQNA8BMf2cFLUxAgbC5w9Wy/IUgihwBDjPE33NiPw
ibUFEriL/0tvnI37AnUFspgDyPfZi8F5ApvjW0vsZuH0BnYGLQYAyK59t2bp8nUL8vgY8gy7dy+1
Bj7OuTiAfQW5NAZqPO9baPyZXvf+UlDnsVEF+gTT3Xie+PDyVoWgDPYw4+PN9NRoDCV2DMq3z3Cx
ZzCyXKOwgQTDoek99n8FacA1TloBQBFmobIXTrce0gfIweEQWQvBqkQk/Hf//wRW6yWLVCQMi/CE
yXQRigoFCzgOdQdGQoA+fYtbLyfvO/IrgDq5CUCKCIUeW7oaddUoXjXrBzoZ+7vt7Ah0BxbzBSoO
9tkbyffRI1fSJ7ZH9fUQHXQxkPYl190MqotdDPi6EA+2OAId/EHXA2ZX/dZZQxxZRvu9wItNBMF1
DTN12GOaQMxtIFLr9kkUm7vE0lldTURVDEOTilbi9tIBhIoIOgIYQULEUNFO4NsBAgorwV1wJHZo
629saQhuiXX4gD8Ao0itQ791zvc+Jg+FMbUkv4BZukYNIyNJRg++BD5/c88XNxFZXA6IRB3cQ0ag
/db+g/sPcuKAZAolyThN3Il/G99i+17cLxAxDImAOB9Moxs590rQdfAXT1oBRlkLlvt9D47OAFRq
FChj+PbtUJOfPV2WIF3diBlBR/vi6xa43CVsCLRno7aIUA0pyH1r2O4+C1SLXfwgK/NQrvRseHkW
emzw8HRRKwPzPwj8G+AcPo00CAP34c8ryzvzG7+1b40IAXMb94V+K4vDKzED7Ru1by+KFDOIrffx
fPXru+7fvvxB/4XAfA8GK95AGQuIEUlIdfdm4VsYBigZUA2ND3lYcJ+5dLae+C0AJuWgY7r3W6Ym
kJFJGmcY/Bv8hQdlJZtWRDcBix0c2QwLzsT701zb6mzBHIJxGAzoKEMy1lHoWSDJgL/927dlMkY8
QVko6XwMPFp/CBvIg+k36x/W2rEGBzCKPxwYwIPoaCj9OwcwweAEnQp8FLppW0kIQ+nZ6IhNCMHw
QyhRTXRBA8NJQ81PwkJLOEbOO96NRBHc8Bdui34hJYoOiAwzRiTrFEjJIc0nOhgr8w7ogwxJMwjo
/Oe2Ujsn/F5tNHSzvbPXBAM8AxLtOMj05QRZOGoGvqTrlZPu30995POlZqWkD4jI+9Ntc65s5BVQ
pM2BWVlfnOpLO3hedBTJahoGWYPADc1+rt/1+YpEFeQdKshQJ6FcyLMlWcjIRd0W3G0IBFaLkdJ8
BIoG6NL/NV4NNDXfiAdHWUZjgCfIl3pmFp1EVi+8aNwlmp+uDrxZj9Dwhfb+zSGdWxUVFFg0dFli
SL4vOcBWXMxTb7AFm/w5Uf/QZyDABrcD6wOIWJRwny3MaJCYhCZBPlvMvW4TSBfYfCZmK23DWX/4
hBX4lU5MEukcGGwMqxmdQ1MdaWJ2yC2jUw6pNJDtxfcAUlNYJAwyQmNmLhAAcPj20HowGd3myVc9
utAae429Q0/f/zgvkn0L1thTDsYEOFwMPGS26htcFXiQ+OxMQpfXIgcbIfaE/v80lZARroQFQULn
wn42HVloeCY6BrCXt/8703xOg/oBfjQEA34aBHU/aRls92x0LmhwB+s9FGxBBnkGaChkZpBBnmAT
XFgSrtlh0NcIzk57LQszhGQROwOYemf8CngZBqNnsxPL81nqAPAK8HVcEEYMPYMBucgA/AzyZomY
ri2NFmZYFHMMAjbdhgIzJDPSDgQ4F5qT7dwknQYGCAp0+KUCN8E0OyLd6wmA+S5+DC41SNEMOMfI
KsuIjLGl3xXtIkI72H0eK628DW+lL/CLyAPY5hTB6QJ8C4PhA9xyAfcD0POkn/c7LkMG9iu0DaOs
rM19gKQzVrhVIt4ucg0Vc4bdtu+ENadGpEYNahAPThjsJsaDxgLaVjN4hxZv+rzJzQ+ewV5YPMSt
4xNLZfxg8OhDBIKbeywKcAVWJHY11Q0c3M99MF/+BDDwb/HW5gVQBesOnEB9Bo10BgHhnmsrCg8G
hTgxuff61hU5DHzLi8aHWFmgoWcqQ9lgnztoW83fqH1rgf7/AF/qA1Xebo0XBtJ0SjZPF0AJfguK
deMv0BMPPkZASnX1yT4u+a0ssRYnnfxmwAKJRfh36lRpAZP7aqUS7772Jf8/C1QSBHym6wvRvrV9
gYp8N/8uqE4Rf/SAJDnYegUcQLoDV3eMrauSARrnMBvYEOUz3p4leNT2sXXoXhuiqQu4KF8cDFg6
RW2Lt1aDPAL0fQcd6RYhDIUCaUVTp7vFf6reFTnvi9hZO3dZfB9LbBcGPABGCgNONsFh4tJtNfgI
BjvHVOBcFyy04PgDOi+9XAOwtdJGFGgDmaVvGfpcw9rctgPKrmFgOkiLQwre0KJgujWcAqm7e7eT
oUNmW+BDEgyDwwYOoGEXrOINCuRDj0PAXu/egold6D5/Yb4kRvp0bxNi3N6r7HRDGFeocexh/Y21
lUVZi4YWvugX5BDYP+xPC7eNwoMgLMYFCfTrkAGOxwATulUPjCJuPHSpAauNX8m/DCN+ridHU1W2
bTPtGIe1HvFVxwFhfdgKLDzhO911PD66dBGNg9uhrxhgzlb9iSg1wpVrJPwhfpvbeLMIEIlsJBR0
ixhROae/rXMLDxhAaFXrAVWb+AVzf9m0JEQQBtU43kTBPGBGXo7bbXfXyCHXXThQVQo8VQZt0A6V
x8RfoED87MzWU0RJZDGOXARVU5/t2CEbVchTV6Zo6IVTvNm67S8oJzQ77g+G2ry0pCYOAkZXg+YP
NmpuG5sDyiEB/lMPa5hb9yAahF+IDX+Zi+1jbvR9ZTr6WYmNJKoVuqUb35IhHAMYEaZ4yd2xEOsE
/OGDvwomWZrObDafDQgPkcLXvDkMAw+Cg70ZVfTHuidGLnYVVtWBx1LHzgA+24sHPRhbBnThCDxA
KE8oxlu3Fo1uwYv9QJJFSPrWQStZdRJWQ7out6G/9hyJrCYGBxibc/w6ITCsiz9iB55B0vbbHiQl
IEfbgxIY2XIhuu0e/w8UChS8Jf7ZU4zwDYuEtsfxU2W6Z6ELkSR5bERhDT/1YjRgSxrVXVuBE65Y
j8R3e2+PK+RcplT5csXi4BJdnZwWEQIQamSM2oYxqEaRfNY9dHMhBwe+uHQX6KVyzeIhc6R6v32b
xdsmDhB1DXQiaKx2i5POKg/MEl/0VnmV64GFHA9t0G9XO2rdWOtxi0PDO/4w7ahweHRhU7uTpk91
SxhySnBRmT5TLpDBXYNHHLSDDmj/LrIQnzp3GNfgU3cjuAOTVWs/oP51pupuE1JCHGC+nKJXtilO
GgPQBTIHVsPrhLhj4oTRAGvIltnqtezE0BwssgU76+8dpL4AQEHTrp7GqsvtFFFC11+GH4228Cte
IYFUhesKG3D3YY13BNJYajWf5NJ2uq6Tolae5oARCuOR3dnokxWjXBEoi0CNVxxwW0kAG7MjHPyM
URVo5D7EWQ0z9KMLqQZcdZsxlQEMEQbUGQ/kXd/XMTAEMfotBWc/DGXwgMhfCVE2qR8tPGyq+FdA
gEej29UDiMBAQEN0Wd5gtSuPdE9EJLPdQQbrXiQPIC+KDmg6SbWC1PYcdRsYyPaRsHXF6xIZzJe4
5bYjRi4RdefliVzm6g1M6E1AdD9pUFVqJQMUbWDvz2DqDAQrQ1k8SvYMC929a0CUM4h2T8GqtcT5
ECsNUDYg3Ub9TsArPjYX9g7ZK5Z1KiODK+3/diQGXCtAdQNLea+AZCsVatBKuIuBvRF7qQHbttU+
PgY9E/g8SxxZPBuwK4C0k71L7nQPLctZQ7XaXuM1K720gLO603vAtl8h60yNPC4oB7g6ige3yWWz
IycheAdT5W4bcT+0TnmxdZG6Njha5HwK3kC0vHAHhgPuzl1Zw++L8VfaGhZaDjCAQif/N8sOjbu7
IIXbkZ2Ed8vCuwYZiANDRww32R8DgCOwO2y4AAwoMhEQPI2Edgkah9V0HMUXxlwZ5CQFOu7mcWug
4TUdEhAnC1Y2mmzUvxTpXE8PiL9t1JRGVbVAXcODJbi9hdpWeGD5bIIFCy7ROBhk7VNBzjkdVmbD
/RKjvAQBOT+jFxYIL+sLTAf/lg1wS+4TPN8cHHu7B69jKn/kEFsoi8u9ES3eKw0UxI2jwIK7zcfa
SYzvKwQPj+a7yBO9wDNww3ciU4vFi89aQxFZkS4Dy8jzvIGdGJTM7pFBvhkGgyp/fhXPtvFu7oC4
SgUJCMd0ZLf3smeRig1h+CEF0XJ724hEILswfAv9OX/FGg4PiojBAwDlIw34W8qHSKEZa8Bkh7+N
frFVFYIMfsE9DDLrn/ztiB0EIFUVBnwJPOsHYQnHZwhGfeEHycN5KJyRal23ALxGLzVdYOsFng9n
BjrDqog5ZrUK+SQR1B6yUd/HwIQ9dNiEqRtURoGwOXzetzDSXZkAEhecX9+4Dj46U7dT/zCpEVDD
S9u3Skc7g0aPOR514zOwyRCyc0srsBEU7w1eLbP43ljr9911Ffmq8nEQQfjCXFdqvAujIMCnvlO7
YjV3Rkeep9ozW6yZHqQU3fCDrEh2c3gSJ7h4r7Y02MDg5EiG4BgzNU3c8PB1qO1eINOdfyaqBmjo
Ks1mJ6GE8FAt0WQyNwitgShG5MjBbiwhagUZlCk2ZJNcTdwzM8NLWMjP9CS49EcwYcWSECZRvq8f
bQ35S0EEPDgWVgalDz7xm8H84ylgMrUIk4VXvRB/Ks9hA0h58OgPA8dBqdYo9t0SPsTusdo4dcjU
vYvHP0UWU7Ng1sKyCpVC8QqQDG2OVQuwoX5N1z02fxKNjWDgdoeN/TJHFNWYgtFt6khjbMyDghcd
fLLELTQKUPboLIs2q4KVGt0bGhatrSx++IPHD1d+adg/LF6IXhbrWVeGgGYIAKsuhgQUjIpO/poJ
e4hGCWRcoXxo9CokxAbrIwYciZBdDnO0hQ/+N5/hgHZhImY1UT6ErmyqoXR3EfkThJ8GxP7POzUz
0jPJ9/YpJXr3I98PKoNBO8p88dx4g8AKMAY9tBd2DDH0EFqKPxdiQGpPNIAx29thQbkxT1n38aKA
qBGOBfUoEwBcya1yyckZ3fwqYsEgy4CAgIFPg6EffIRZWWd11BRyyUIDqwhyCAribR806NPGA6Em
fata6zzb7M76IjlYXLb+hRtPO/PAi1ZYO1BYc2rwwj+89dJR5oH5/H9camBToNxB2EIude9KKh0l
o1MToHonH0KwrvOIEPOzWIle2501vFx/momuQHi2ORWzD+B/dbFXjX4Ix0Zc/h8wk2N37v92BDNb
QOFZTxRXc6/OdWkUSmlfZ/z00R6Jn4RJMFP/QFzorKGNr1U5zWFZnA5Rs2Mj8agDVRcbSVkyBinc
SZXoNPpQhIWGgfGYOcfOL8gJr0pWz7AJ3Y4WdkZKLRVZYypXdWYb3FKRzohXwqNvSG1qpyu67OKK
BEh05oatu6Jftle/0Bz0Ldy14plDD1bGQAH316D7VHhZCQIIIwB2ByYUiY9M8C6gjG6P1IJrRHFE
gH4sdSCjbhTO6iscYLno9PBScUdkSAWFKD0gHBrf2MjOrf4R6xiLDg04ZdSWGQ8KfHW40wm+YAcE
DINkJDz9LSL2K6LHBYVL9q8Q5usXaOWkUTnHBCiFhgfeOA9GfUvgYxQr8Bc6AQ+U2CHQsOGINHB0
7aCJ32hv38l0TkOAeER1D0VweopOCTq4wvbnSAl+SAQ7TB5y+QW3A25qh4TXgfvsfB1JNMcGeEsm
gf2SfhB9vc2VGHMGXlkIrCSwQUttFDvFTfNJWx22nzIEcyiNRhhNHlYBJ03uaOta5RisFronmDT0
Eb3pYbPgDrIdcQ0EUMdkYIPHHARog/sDk+IuCAs4Kb7bZx8Auw3gPXAXCsoiSGa+3xZ7VjqNo/aj
0ATUTLrqa8PBgDOgQm0IPmV9DDd+FvQ8Fm3hD7YJiVFaAogIturERoDtLlEMB7BFAWWujLHtqP/2
vwgsIVuJXfg73n9mLcYrrVAhGh0MIcvGR27Ad/xjMqNJ/zeLtKK3UrhcHBkEA8a6uXdHs4sHHjvY
dCNxEytVrtsNNHDLDDMDSSvW2Gyt3f4JihmIGEBBe/eLYitbATtHpgtoi18OPHR1iSNcdwVeD450
tYTtw1KbHFYaBh4zHSkLNMrd/FYINIUD8SFCg8HCF1teB1tLCLCZjTjSfULWS7m7Uz1EjV8BWYIe
hbemi//Ds4Vaz34TDhfcQqVEt4uQ7m4FSS7UiBvCf+24CX0j31pn3xkUMIC6GBZDg3zt6w5brZp0
FDG1wMi5Ff7/fO6NUQM70H1lO899YTvBYU9cBu9aG2y7IUgST+I7wn5DkuEd/DvHfj8rwYz/B3w2
LTnmFhv9A847132jAZEV+LViF/BCQYH6BHLp9iENPOgQDoMADtVc+Iv7O30WjDFeBEw9lMfzuBAA
dXwPF1DOAnIDbD8s4ESAT27wD4SVpokMkwDnavgShr5FK1NRv/0Ob2+GW4sqcldRKgL0UOsWWvjQ
Tj3Mc1N1+CIFTcB78Ru+Bh/jXLysAY4OTdDNaOM32ij024F9+ACw3Xf2Bcy6JlMwV/BTrgHXqqi4
+aYOiNWBSRZfhFlXJiO/lMxWzW08mFx8Hq5ktgjNs8/P/sboHTRrjeYCMwDCDPCQZZBtaPscYJ6z
BN/DBFckBP+8+41b4Tv7rWRb6+xHZItPYDEW29h+dlWJTXA2bDpwhMpd5WDV4IRNaAfx/C/cSvpO
RHPBFD6IVAXgOBw+ulu1AMZGIXLoPwwc/A/DMbmDRXBE/01sgrYgm9lw/PxgCWTD1m5Mc+sItYHu
CfNQEwhdrVjQWEL9RahowC3s+4QaBKIe8KiBcoleL3VRaeqo/iZUoQKS6IRqZ6GZqACTQnAJNYuo
hQUMf28HPU+TWZqb4n1BkMhXow034P4zSIN+ICgPgrNZlMn/OEsftNRGLHA9+xFwBsC7QKMsD3TI
QAkCbrC0i+hhfe9l6Jekg+8tRDEtag/m6Amt+ETlNBFMfeh9Wru9RAYAIAM3DYFjtxu4Yin7h0ct
5FCMamcvaFy/fODXPW3X+wwxQAEeUsckdaMr0SNbRSQumTmy7zHILT8cGa455EgOFJQMDMnYC3R+
FQRoPttAjvwtngnAEgtJHdv+SR70LbcU/DZ45/DMw1Pj7C1wBsycAkpEk/iboiYfOUYgdzXrCzKM
0OAU7JytdVhxoQT0G3UKGIbJXetOxMEPAnUJ2E92BKdfdFhcAgxXbC7YxX4Mmjv+N0ASOWCmcI5k
Wzk1zBjdwTeLHVxE5DpN9Zrf0wmy5NbCVLMmmqQZNqOTapQVehHlGCc5MC5oQLSk/bPNQZJWk5L8
FYo8Ee9QdSM1ESTGE2a7kHUDI9TrEcju1wkwIKisNb3QPO/cbBuEGwjRAHSuEZsZRpYJ0pwPWsXZ
N8omUL5UUCtM+LEvE/alEHQgaksoy65hHbhIIghTCOmJ2CB0BqcntdT00Fhs6UPN9hm8OMhD8T3k
WxApHwhJIja3hXz/UC7SR0Ue8rxoQC49eIOng69hvoRMu7BWRf3hGSAJU5QUZ7QO88EeLDw0Sbzm
s1RlKPj9YSVskJdQF/j9ChkANpzjU6ZNYBfNlh3moi3XHLJMDOGRGWoFDgcqs4GDpNNWrCpQwuLP
6YpgAZtWvhEB2N4T1IqdDRP9daR7yeou4CVpD2erEBvGDmfd/ChWdLMyHisw9NmMNxqYBiJooB/l
QPsrxE5Z/g8aBVp8t6s82ejdGVChav/bUAAR8ssNoiNUpFWVaACA0MKQS9YK+gPwIlJ/kJQWPnAL
Cwi5J/fWAbX9l7oB58dTwU6L2PfbjTzfiS/0l7ofihpIM94j2cHvBDSdcGQZa3fdM/dCFBLuPNsg
suf+3yUSSK46w0JEX7LDW4TAj/z+FooCM8YjwSEEhfBCT3XqDoTiCx730F5d/kzfb+EAbiDwzwdy
CAfaxM0NxAd23vDUBwFyByddYQnlRRP29mMp05Ef9gpVwU3E2dpGcMDElwskBQWtoxJ99maJAQ2q
/A84R9+XBvpm0ekYwbsadumcBA0IaldWAB16GqEYSKQ9A+z61BZau5DrHUp0MXXxgF7Y0LX4hol2
dotWbGB4eAOXe7wZ3kJ6dctoCRvKUSfKHKFPvXxzYL+AcR1orAFZ6KBW08nammpr+K79W8YH9SyD
bK7AJAJADJ7l9qg6Jn300f5sTVUK4LIek7g5ZDsIL2ouC4gWS8QWZNgJxNlQrjRs4ksDBG3CUEa8
BTVNt5mOwb4DkMCSFrlW2C9XaUYl97uh9nXdlArEB5YX7LxdzW3LwgkwxgKY8beoba6h02bKCAWc
C22LQSX8vw3OEG1C15WgOtIDpDeD5osFba1QgnjUa+65tqYCshYePDAFKMQMFWQNVBDB0VvmHma7
WzDPwrOfHzuHhISsNRFrqlAxBwEmadNwgNgZYaX4neNkIRv4wD6y6LyCwVQxLTI89my4LB2IAQIS
jBSsCLHCTNGuypmiu2ytV0U12AUGL9xnQ9vdywEuB94rWF3gASucbM/iAexr5NiSqOgQoTcE8j+W
EXlO+8ZeOgD/lAMTBVdDagZTstEjZi+59upO4MAc4WaEZupQgfs4ZHPu6fjP9Gh+ZgSAVuYRTAWf
aDfb6xgNUD1HJy88Gmoktu6sMqJq3Agr11RVlHL/dNjraz0zI3BXlIWiG7b9Qm8Dx74G7A1GAZSJ
nQwA01BsIPTdndYBXzBRRT/+OjezhocIwWiCKUFS9uBkEHQYsbCc6IAWEwliEQx/J8wlFBAKkWhw
MggJTFISWYcEpyoYYSj9YtekwghmgmoI4GY/G0pam1l07UnJ3CL2ZuTkm5NEEbAJDsDlIIvmN6t3
67uGoYds/9hiQZKYx427kwVbHfzVU7D0eHKrZiv/XBHhanhgGBwU2gUCLTiAhbwMoI9QpmNVVxT0
Rmo/RAsbC9HyXqCNd1AOUHuy4FLhtGtoTnXlRxdqhJ9FW7ApU4cIg4cVFOrDBFZixmToJsQ3g/pi
fUcqlDyKS8CshLV+MK3V28iBHxw7ytMjRGUrmkH1fQ3vyT41iFyJWFdaAzP/XP+b7PaL8gPx1n4Z
FxoVgMJhiBQ7/c3VrUewfOc48TQHxkYEQDYuBY8jg+ADZ/80DxOOckEWyFbBieTLPrLYuAh9QnEF
M/a9sht8+oPHA4B+HXKUM2///g8CRjv3fOOApB4LAF/rYDawHkbFuwjDuaiv28EIA/DE0rBNAHXy
P0P++t+2b0PARrEeH8nNO/J9DIoMxbAy0ttihHDr/MU7Fre7FYB2tsWsC42DWyVLN4yFXzL4ueSB
XDIAM/iLNJ8B/LOkVmsE3b01kIHDtwdoXDQIYaziH8AYNgZADmQFDwRyu2RABAzWKDOAHMhUDDCQ
5yG8OzYsMwTa20cWtDJ8FgRVfRboZPfU/SVqAeUsfBIVfA2OgDPdEzD2LQwDmdncR1eInrQcBbVW
j/02HkB9e4YeATgldSGNbLMi14a3UGE0tqlIhMu4UIBtbLm0YPO19Py/IFc8ByN6n7aInRMr9Pzs
3aw0+Uw/UIgYUziRLcDwaIijyEQrGjvbOBgpzxxX1CbPEDatKLXsxS70BnKkAGSLQTs34MH8Elhg
IGbPznNzAYQnaIB/aEqIMyMMUPzDIJ+MjfgPhCIZYBEhDLdDvrxVVE48GDxHB64/gf9bFMKZjbTy
C+z2K4gAKOFiTYJ80bAaPnE9HAnFzBJiBQP1t490FX4M9wJ/B2h8NK9Wrn0C3usFLg1DZ4clSAlG
B0m4hHVEkS3K7Vz4t7MzAxsrYiFKdA9odDSs1Tehs2YcNw59h+IZaA2fDmSMH7OBdggTvDgneMKM
cHQJPYi2WycaOiOIMLgUh9hiB8BeuPBqKAPQ5oVoIcXUqAUAADJy29CENSBN4AnkIOg0zmXz7Mg0
dfD0jClJin5hDDvWfWnIwVPJBIpuxoH2R5pePclFPCByODw93AD/S/w8K3QwPHksPH90KDyAdCTD
ilovASCIBPgwn7rbk0YKxhUNRgQK8buAoG4B2yQe/0YBzkfEVipQ9+znYwixfElLB/Xn/zPJQfom
/lu6yn0Ji3TF2EBl8YN8xdAECbhN3BHUU8YH6M0gEEQQvpA1cr9QNOi886WB/aSKTA28jeJC8V+I
CopxcAEH/y3V6sHhBD/QzheISgGKSJZlWboBGAIPAgZe0O23zxkCikAV4D+KRAUMQgN1pp4n9RgE
V1gCBcgWPCLT3ylovDoYNehPZNYEiK31RfHsMATwN7pQlPLOciI77Fec0YA06Og4OYAmt0U5ZDHC
Rvp/L+GzLoqEBSeIRDXzdb+NVSVqG7oZ9CRjYlgMXYhab6k1+IiQkfCDqHMvvF5Mcg1hAw1DaQcK
A7r2hQ3+BHLZpjJX1diFrw03mQmFdCpN+Gy/C2hzBMZF+z0IAvo918StARR1HzwD3qUMmlQqOKK1
pJhauEEmBxRRUxTYpk3FhVOzQPG7wMOykXAQl99QBXvhM8YJD1JqLpg2SgTQdK9meFctC3BWGvrI
WFktJI1DBBnVlc52AKogaBiucSAS88UbHCcQsgaVFq1ZtdnIvlMbUDIMftlCdtkOMK9oPCARGIO9
VAuiGGgImjWUHdm3wJQUaPg1M9wRUk3EyNTVOVldIbSgcwDRJwAScrDUuDdwyIVY3v5zWDeDyh12
9k5QF1CEHDLLjbpgP3UD3q5iUUzk2Yx4SCxEuDbZCDQ3dkfGUE/YDbCNnQhShYvDdk1zCYpjxgUT
Zmik9EBqwP8MHUgEOtGNWe7XO/Md+QYxoab3Bw+Mv2/ID6hIBrj7DI34vVPDBRFc2kTkk+1mFA1d
mwpe0o21oe6oEWUSc4uFov308YbJweACRrk0BZ8j0Ba2WIoTCtdA2FmJh3RgQHQeGE2J7zc7ZNkK
cmX54CdMTzIWdW79AW85XfitIssDavjswxElSGAmdfiuOoc/FAxGVzl1ELg16gURfnKLEUQpfUJH
banJFIz5TSSYVQ/q0omDwtWAt1sB7Axp0g1w9XOLOlK87P6JVfQIZeph2X4m+Vh915fMEVp0FIoH
Fkc8CnQK7mrB34cDxztFEHyXpS+IHAiyVPsRn4PI/+v2N/5Yv4GGKMMJOxeAPzB0GW7ksIhXEAcw
HwqWCANQpV7LLfxCkcA78FfZYw6zR5aRbQgIWgxREA/foPvNjkiKBjwNdAyOCBJ0BDwJMFuB+HUD
RuvrdCYqiK1AJKPIJUbumu4X4T48OnQ5LjUxKgIEFxR/W4rsDzh1CTiEDf9A23XQLhADBEnOiBDR
d8Rd7kGB+bZyvusBTkVibKwlEgBdzJgsz4XID7gA/9Mgi7VdzA8OJDgrHC/D3gyQ6Tg6dWEeMJnh
RP5bD+igZ+5ItkBG0soBRulcB7vO0k/1FsG5YYK/gaFdbeIKQjvXfOp13cdWEGUCKkIdC+M37ilq
8D4KqI4qCXPtN4gIgg11DusLIAsc0NIQGwcGNQ2EggQOyEudj21rBBeGTornHQUEG2wrbTADhkkA
jpI1M8Jyw2MNdYTzqwybYJIAGI0bx4UYMJ16BU0GtmgxomBl4xEOZ+MG01BRUGT8m5YQ/YK4i8HH
aCthor7aLBQ3Kxpp+wAQ6g+IXsKAww/7iB9wB8VWvtoziuW7314XaooRgPogyvoJdRNB/qVSbwc5
fxK33ASAQY1EQtDNGvH/HjB96YA5LXUceU3PreAQVrNn1X9uSVGqs7VWYt4QDHLcVYBoRDhKSDey
i61oqD0b+/agF3JAIYpaPTQEhmo9EAd+SDSCLrht9kBTaHWSj1T8agYbmak9hBnYg2DqLQIXLzj1
V9SPD9w85foe8r6YOvjGHzCYXXVqVOiIVlMpnIt+EKa+RJWFmH3qcozEPZB4jbnc6LEkPwo0OIm/
ECfLNmvO6v5XRUAYfEIy2O4HPSs2fjw4KPk838ozdE8rj0Qj5MAuFDv9A7nkkhMIBKckj5D71wDE
55nMwWj8viEMtXp8mZGPqt09Xc2S6TfA+IoBi9lKPBUHDlJT6UOKAz9rAxcDQxXgG187y3QuUC51
EWrNai+ASKG0RECscVsMwxIrwfwP8u6t0FxOwhPL66woBWj0N5kzvAigtwuStaVGeHwjnX2/7Cao
UC25H4gT8xJ0c0dT6wYJBkZTS0PDKHXGprU0A/IsNOAi3FhcDgFJuv8QTCIwNgHYQv9sL1fBIBIC
b5cPqSzVb0UREAzc/C1QKTohtVdZI3LwICVTS0tEDQkgb3C6E4c7grEZ/d5WTAK57EhQFtQJmB23
o1C9DSpIT4y9HAF9UzxUc3vgdCtqGRthCrKJ3AhD3nOLcFSUA2tDxtrL1Qdvk95LAE4Me4zp9HUY
unVwQabqndNK0wKuDQMk8CcYOCSWgnxfcgMBWw2viA0+ZuxzAOnB+QNR6uz8GAEL5Oz8AIIVn4ZI
XEBXblYgdtGE1es1wePNJSNP8HQk7AzuP4iXLOx0IpvHIaYeXQDQPAO+p+IG+vgJD4et3ySFRHKL
fLMNnHE7aXD+FIftDrJwtmjYx+tuDdCHPIc8YMhSwIc8hzxEuDashzyHPCigGpgOM4c8DJCJ1mMm
3hs76weApQ07BnRKBoTYVY0IDTvIArOwxhBosg9TcBR8vqD2GmJs5z4ZfRFHFW35PtE03XZAFBSA
ZCkDN0XTNE3TU2FvfYubke9Nmf8lVBEFCBDMzF8gDMRRPXA5CHIUge2P/b7pCy0EhQEXc+wryIvE
DL0uVeqL4YtTnFDDkgoZRJEAqlSpKg5ZqopCgwM2zUFRqBwBQ6Wil4ibdGVGcLe2UfRNYXBwwEET
DW5kC/YMRYgVDgNeqBp2cnMPd0VudlF1FN0Qb27HVrd3h3V9YhhXK293c0QdZWOC/Xb2dG9yeRVE
InZlVHlwJHbvZ/9HU2l6ZVpDbG9zChRUaTX3bt9RVG9TeWplbQstHBvbbkH2QWwGYzpUGNqT729w
KU5hbUxTUG9HJeyZqJIhPdrW7b4OQ3VycqVUaOdkEVeJxn67ze0KTG8QTGlicmGlbF479t41cmNw
CY9IYZgkcNvawa1BdB0qdTpzQbJbsIEyNwhuQZ1ACNhtUBtoQYkKW5612GQfHkxhRZx7usNaGVFN
X3hvhzZZO1hdRGUGalOLQGj/VkdNb2R1FRQYwoTYd0tVu112SBpBcxhTCGVwBtiWS3hFeGklYUaY
U+0w9+YOHE9iasCkULDfsCW0Y3kGMv1pgs0K22Nru3VsTCm1UNXNGmlaTUlmgNpF+W1h5RcD4/2O
cFZpZXdPZosAYgkrtEw487kRClBvzA1hZGVD2L/ZW9smTfZIQnl0Im5BZG7CEt5kcnIWx61uWWu0
SKU4HCsnw5gxexMZYAS8rDCEbqrNCWlBd4+zYY1GSXE1a2VkE3ZqC6VjEgsVSdKZYZJuUiLkVTM2
wbCw9dRCkyZLHYUUnHmitdqxx/g2Z4xLZXkMT3BN3Tr36AtFJA46Vo11ZWEHAIYPJBEJM3cppnVt
MAyvrdlssz9kwggBbaPutDXMc2WiandDEPPY3wwDB2lzZGlnaRl1cHBzzc22EXgSCWZbCDjNVvhz
cGFLT80sWMD+e5tVL0J1ZmZBDwtn2o48TG93d3Y5crYjUZht2HcKR9gsy7I91BMCCgRvl7Isy7IL
NBcSENWyLMsDDwkUcx/IPxZCUEUAAEwBAuAAD3XLSf4BCwEHAAB8UUAQA5Bhs272DUoLGwQeB+tm
S7YzoAYoEAfyEngDBqvYg4FALs94kPAB1zWQdWSETy41dCt22bLJe+sAINULtlHg4C7BxwCb+7t3
Yd8jfidAAhvUhQCgUH0N0+UAAAAAAAAAkP8AAAAAAAAAAAAAAAAAYL4AcEoAjb4AoP//V4PN/+sQ
kJCQkJCQigZGiAdHAdt1B4seg+78Edty7bgBAAAAAdt1B4seg+78EdsRwAHbc+91CYseg+78Edtz
5DHJg+gDcg3B4AiKBkaD8P90dInFAdt1B4seg+78EdsRyQHbdQeLHoPu/BHbEcl1IEEB23UHix6D
7vwR2xHJAdtz73UJix6D7vwR23Pkg8ECgf0A8///g9EBjRQvg/38dg+KAkKIB0dJdffpY////5CL
AoPCBIkHg8cEg+kEd/EBz+lM////Xon3uQ0BAACKB0cs6DwBd/eAPwF18osHil8EZsHoCMHAEIbE
KfiA6+gB8IkHg8cFidji2Y2+AJAAAIsHCcB0RYtfBI2EMOixAAAB81CDxwj/lmCyAACVigdHCMB0
3In5eQcPtwdHUEe5V0jyrlX/lmSyAAAJwHQHiQODwwTr2P+WaLIAAGHplID//wAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAIAAwAAACAAAIAOAAAAYAAAgAAAAAAAAAAAAAAAAAAAAQABAAAAOAAAgAAA
AAAAAAAAAAAAAAAAAQAJBAAAUAAAAKjAAAAoAQAAAAAAAAAAAAAAAAAAAAAAAAAAAAABAAAAoAAA
gHgAAIAAAAAAAAAAAAAAAAAAAAEACQQAAJAAAADUwQAAFAAAAAAAAAAAAAAAAQAwALCQAAAoAAAA
EAAAACAAAAABAAQAAAAAAMAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAgAAAgAAAAICAAIAAAACA
AIAAgIAAAICAgADAwMAAAAD/AAD/AAAA//8A/wAAAP8A/wD//wAA////AAAAiIiIAAAAAAiHd3d4
gAAAeP//iIdwAAB494///3gAAHj/////eAAAePd3eP94AAB4/////3gAAHj3d3j/eAAAeP////94
AAB493eP/3gAAHj/////eAAAeP////94AAB4f39/f3gAAIdzh4eHgAAAB7M7e3eAAAAAAAAAgAAA
8D8AAOAHAADABwAAwAMAAMADAADAAwAAwAMAAMADAADAAwAAwAMAAMADAADAAwAAwAMAAMAHAADg
BwAA/98AANiRAAAAAAEAAQAQEBAAAQAEACgBAAABAAAAAAAAAAAAAAAAAJDCAABgwgAAAAAAAAAA
AAAAAAAAncIAAHDCAAAAAAAAAAAAAAAAAACqwgAAeMIAAAAAAAAAAAAAAAAAALXCAACAwgAAAAAA
AAAAAAAAAAAAwMIAAIjCAAAAAAAAAAAAAAAAAAAAAAAAAAAAAMrCAADYwgAA6MIAAAAAAAD2wgAA
AAAAAATDAAAAAAAADMMAAAAAAABzAACAAAAAAEtFUk5FTDMyLkRMTABBRFZBUEkzMi5kbGwATVNW
Q1JULmRsbABVU0VSMzIuZGxsAFdTMl8zMi5kbGwAAExvYWRMaWJyYXJ5QQAAR2V0UHJvY0FkZHJl
c3MAAEV4aXRQcm9jZXNzAAAAUmVnQ2xvc2VLZXkAAABtZW1zZXQAAHdzcHJpbnRmQQAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAUEsB
AhQACgAAAAAAgU0+MMonH54AWAAAAFgAAAsAAAAAAAAAAAAgAAAAAAAAAG1lc3NhZ2Uuc2NyUEsF
BgAAAAABAAEAOQAAAClYAAAAAA==

------=_NextPart_000_0014_F3116968.04FBB8B4--


From verhoeff2 at gis.a-star.edu.sg  Fri Jan 30 04:48:17 2004
From: verhoeff2 at gis.a-star.edu.sg (VERHOEF Frans)
Date: Fri Jan 30 04:57:23 2004
Subject: [Biojava-dev] BLAST(N) version 2.2.6 not supported
Message-ID: <6D9E9B9DF347EF4385F6271C64FB8D560B06DD@BIONIC.biopolis.one-north.com>

> By default, the blast parser only handles output from versions which
> are definitely known to be compatible with the parsing code.  In
practice,
> you can often get unsupported versions working by calling the
> BlastLikeSAXParser.setModeLazy() method.
> 
> (Note: I haven't yet used 2.2.6 myself.  Let me know if it works,
> and we can add it to the whitelist).

I have. So far I have not encountered any parsing problems.

Frans


From ANTIGEN_POSTOFFICE2 at Cellomics.com  Fri Jan 30 17:11:06 2004
From: ANTIGEN_POSTOFFICE2 at Cellomics.com (ANTIGEN_POSTOFFICE2)
Date: Fri Jan 30 17:17:24 2004
Subject: [Biojava-dev] Antigen found VIRUS= Win32.Mydoom.A (CA(Vet),
	Norman) worm
Message-ID: <141AC65394B43948BA3701D22CB284A2A63E54@postoffice2.cellomics.com>

Antigen for Exchange found readme.bat infected with VIRUS= Win32.Mydoom.A
(CA(Vet),Norman) worm.  The message is currently Purged.  

The message, "hi", was sent from: 

	  biojava-dev@biojava.org ( biojava-dev@biojava.org)

to: 

	jimmy@cellomics.com  (jimmy@cellomics.com )
This email and any files transmitted with it are confidential and intended solely for the use of the individual or entity to whom they are addressed. If you have received this email in error please notify the system manager. This message contains confidential information and is intended only for the individual named. If you are not the named addressee you should not disseminate, distribute or copy this e-mail.
From MAILER-DAEMON at acsu.buffalo.edu  Fri Jan 30 20:25:55 2004
From: MAILER-DAEMON at acsu.buffalo.edu (MAILER-DAEMON@acsu.buffalo.edu)
Date: Fri Jan 30 20:32:19 2004
Subject: [Biojava-dev] failure notice
Message-ID: <200401310132.i0V1WAEs018604@portal.open-bio.org>

Hi. This is the qmail-send program at acsu.buffalo.edu.
I'm afraid I wasn't able to deliver your message to the following addresses.
This is a permanent error; I've given up. Sorry it didn't work out.

<james@acsu.buffalo.edu>:
Sorry, no mailbox here by that name. (#5.1.1)

--- Below this line is a copy of the message.

Return-Path: <biojava-dev@biojava.org>
Received: (qmail 27174 invoked from network); 31 Jan 2004 01:25:53 -0000
Received: from unknown (HELO biojava.org) (202.127.20.131)
  by front.acsu.buffalo.edu with SMTP; 31 Jan 2004 01:25:53 -0000
From: biojava-dev@biojava.org
To: james@acsu.buffalo.edu
Subject: Server Report
Date: Sat, 31 Jan 2004 09:26:19 +0800
MIME-Version: 1.0
Content-Type: multipart/mixed;
	boundary="----=_NextPart_000_0008_18260C28.18A8E52D"
X-Priority: 3
X-MSMail-Priority: Normal

This is a multi-part message in MIME format.

------=_NextPart_000_0008_18260C28.18A8E52D
Content-Type: text/plain;
	charset="Windows-1252"
Content-Transfer-Encoding: 7bit

��ݠ��ogCf�ə/�����3؄YoX%�6�ϯ���B)q�Y��5nd���]�3�c[�L~����s&�ݵT�zݖ����<QH,}��3���"�
x�F3���!�zeX*s?Z���v'A����Yn��.̲�r�R�(TL��e!ӎ���ʢ�*�������U�ױ]�R��I���a�N8�F���e҂дX�>lJB��sM[�F�5C�(��ig� ��[xřu/ݥ#����Z
�'�%>�oZ�) h�G��(�I{��
��3�q��{n4>��2��I�:���h�6q۫���Y�_����u�-�
�sV6К�Z
���7��.
w���wJ�e�Ȥ�hr��}Vf���,�Q��S�Oy�p�&���8�7[ƶ���\0�ϼc��$�K���}����M������
�{*r�,�
�)cVW�&9���|�s�
�����k��g�֮����1jE�u^H�^����o�G�&\�Fc�.A�
X������"YzVMF�����
_�!e��l���i~��L˦�_���PDE9Ŋ����Ln��B��>�W
�[���s�j�*�⬅�����U�"�!yr�㚴*�r�~
d��iW�Ѯ'�Z�?�_��$�8��_� �a��"��)��Z(�upgX3�Y�ܢ��]J`ꦑ�C�r�/W�`�9P�u�Q[v�հ�J�0�U9�u�Զa<h�m�gBԔB���s�


------=_NextPart_000_0008_18260C28.18A8E52D
Content-Type: application/octet-stream;
	name="document.zip"
Content-Transfer-Encoding: base64
Content-Disposition: attachment;
	filename="document.zip"
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==

------=_NextPart_000_0008_18260C28.18A8E52D--


From autobuilder at derkholm.net  Sat Jan 31 00:19:57 2004
From: autobuilder at derkholm.net (autobuilder@derkholm.net)
Date: Sat Jan 31 00:27:47 2004
Subject: [Biojava-dev] [biojava-builder] BioJava nightly build report
Message-ID: <11707096.1075526398225.JavaMail.thomas@firechild.derkholm.net>

BioJava automatic build system, run 20040131

Binary build: OK
Javadocs build: OK
Core test suite: OK

A snapshot release has been made at:

      http://www.derkholm.net/autobuild/

The following files were modified in the last 24 hours:

  * biojava-live/src/org/biojava/bio/program/gff/GFFComparator.java
  * biojava-live/src/org/biojava/bio/program/gff/GFFRecord.java

A patch file reflecting these changes is available from

      http://www.derkholm.net/autobuild/patches/

-- 
BioJava Autobuilder, maintained by Thomas Down
If you notice any problems, contact autobuilder@derkholm.net

From seluja at datect.com  Tue Jan 27 16:00:44 2004
From: seluja at datect.com (Gustavo Seluja)
Date: Wed Feb  4 22:38:02 2004
Subject: [Biojava-dev] CE trace editor
Message-ID: <002d01c3e518$a180bb50$0700a8c0@GNOSIS>

Hi,

I am looking for a Java tool that can display trace files (electropherograms) of ABI files.  How do I search your site for such?  Do you know of anything like it?

Thanks!

Gustavo

-------------------------------------------------
Gustavo Seluja, MSc
Project Manager & Principal Investigator
Datect LLC
2935 Rodeo Park Drive East
Santa Fe, NM 87505
Phone: 505-995-4488
Fax: 505-995-4490
Email: seluja@datect.com
http://www.datect.com