From biopython at maubp.freeserve.co.uk  Thu Jun 17 13:58:03 2010
From: biopython at maubp.freeserve.co.uk (Peter)
Date: Thu, 17 Jun 2010 18:58:03 +0100
Subject: [emboss-dev] Quality scores in union and splitter output (was:
	partial GenBank)
Message-ID: <AANLkTilOfcdvpoQQdmXF5YPeNWRXTxSnpfiEHO3WKnZu@mail.gmail.com>

Hi Peter R et al,

This email was prompted by a discussion on the main EMBOSS list:

On Thu, Jun 17, 2010 at 9:38 AM, Peter Rice <pmr at ebi.ac.uk> wrote:
>
> On 16/06/2010 13:55, Roi Brodo wrote:
>>
>> After some more reading I think I can do it using union. The problem is
>> that after I create the list (of the two ranges) using yank, union dies on
>> "union terminated: Bad value for '-sequence' and no prompt". Why is that?
>> Shouldn't i use a yank file?
>
> Yes, yank and union is the correct approach.
>
> The output of yank is a list file, so the input to union should be @filename
> to read a list of sequence addresses from the file.
>
> If you just give the filename it assumes it is sequences (perhaps a fasta
> file of sequences to be joined).
>
> We will add this to our feature requests - it should be possible to make
> seqret handle ranges from circular sequences. This will be after the next
> release as it requires rewriting the way several library functions work to
> allow the circular range.

This is an interesting example and I couldn't resist working out how I
would solve it with Biopython - something like this if anyone cares:

from Bio import SeeqIO
old = SeqIO.read("input.gbk", "gb")
new = old[800000:] + old[:100000]
SeqIO.write(new, "output.gbk", "gb")

We already have several unit tests for Biopython which check some
functionality against an EMBOSS tool. I should probably try using
EMBOSS yank and union to verify our record slicing and addition...
As part of this it occurred to me to try union on FASTQ files and I
found it does not handle the quality scores properly:

$ more example.fastq
@EAS54_6_R1_2_1_413_324
CCCTTCTTGTCTTCAGCGTTTCTCC
+
;;3;;;;;;;;;;;;7;;;;;;;88
@EAS54_6_R1_2_1_540_792
TTGGCAGGCCAAGGCCGATGGATCA
+
;;;;;;;;;;;7;;;;;-;;;3;83
@EAS54_6_R1_2_1_443_348
GTTGCTTCTGGCGTGGGTGGGGGGG
+
;;;;;;;;;;;9;7;;.7;393333


$ union --version
EMBOSS:6.2.0

$ union -sequence example.fastq -sformat fastq-sanger -osformat
fastq-sanger -stdout -auto
@EAS54_6_R1_2_1_413_324
CCCTTCTTGTCTTCAGCGTTTCTCCTTGGCAGGCCAAGGCCGATGGATCAGTTGCTTCTGGCGTGGGTGGGGGGG
+
;;3;;;;;;;;;;;;7;;;;;;;88!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

The output is well formed, and has the correct quality scores
from the first entry, but has failed to include the quality scores
of the subsequence sequences (defaulting to PHRED zero,
the exclamation mark).

There is a similar issue with splitter which seems to default
to PHRED one, the double quote, for all entries:

$ splitter -size 5 -sequence example.fastq -sformat fastq-sanger
-osformat fastq-sanger -stdout -auto
@EAS54_6_R1_2_1_413_324_1-5
CCCTT
+
"""""
@EAS54_6_R1_2_1_413_324_6-10
CTTGT
+
"""""
...
[cut]

Now admittedly neither of these operations seem to be very
natural for short read data - although it might make sense
to take a FASTQ contig and shred it using splitter for
feeding into another assembly tool?

Anyway, I thought I should report these issues.

Regards,

Peter C.

From pmr at ebi.ac.uk  Thu Jun 17 17:01:16 2010
From: pmr at ebi.ac.uk (Peter Rice)
Date: Thu, 17 Jun 2010 22:01:16 +0100
Subject: [emboss-dev] Quality scores in union and splitter output (was:
	partial GenBank)
In-Reply-To: <AANLkTilOfcdvpoQQdmXF5YPeNWRXTxSnpfiEHO3WKnZu@mail.gmail.com>
References: <AANLkTilOfcdvpoQQdmXF5YPeNWRXTxSnpfiEHO3WKnZu@mail.gmail.com>
Message-ID: <4C1A8D1C.1090901@ebi.ac.uk>

On 17/06/2010 18:58, Peter C. wrote:
> Hi Peter R et al,
>
> We already have several unit tests for Biopython which check some
> functionality against an EMBOSS tool. I should probably try using
> EMBOSS yank and union to verify our record slicing and addition...
> As part of this it occurred to me to try union on FASTQ files and I
> found it does not handle the quality scores properly:

Thanks. Just in time to fix those for the release.

regards,

Peter R.

From biopython at maubp.freeserve.co.uk  Thu Jun 17 17:23:04 2010
From: biopython at maubp.freeserve.co.uk (Peter)
Date: Thu, 17 Jun 2010 22:23:04 +0100
Subject: [emboss-dev] Quality scores in union and splitter output (was:
	partial GenBank)
In-Reply-To: <4C1A8D1C.1090901@ebi.ac.uk>
References: <AANLkTilOfcdvpoQQdmXF5YPeNWRXTxSnpfiEHO3WKnZu@mail.gmail.com>
	<4C1A8D1C.1090901@ebi.ac.uk>
Message-ID: <AANLkTinJKfekPvNtT-ryKL1B5uVucssMSFYCj1R6hwMj@mail.gmail.com>

On Thu, Jun 17, 2010 at 10:01 PM, Peter Rice <pmr at ebi.ac.uk> wrote:
>
> Thanks. Just in time to fix those for the release.
>
> regards,
>
> Peter R.

Great - that was quick work :)

Peter C.

From uludag at ebi.ac.uk  Fri Jun 25 06:49:34 2010
From: uludag at ebi.ac.uk (Mahmut Uludag)
Date: Fri, 25 Jun 2010 11:49:34 +0100
Subject: [emboss-dev] ajSeqNewRange, sequence accuracy values
Message-ID: <1277462975.7755.25.camel@emboss1.ebi.ac.uk>

Hi,

Working on a problem I need to copy sequence accuracy values when a new
sequence object is defined using part of a larger original sequence.

Since ajSeqNewRange functions in ajseq.c doesn't have sequence object as
input arguments (but the sequence string) I either need to define a new
ajSeqNewRange function with the source sequence object as one of the
inputs or I should define a new function similar to the following that
can be called after calling the ajSeqNewRange function.

/* @func ajSeqAssignAccuracy
************************************************
**
** Copies accuracy values from src sequence to dest sequence.
** Assumes dest sequence was a subset of the src sequence.
**
** @param [r] dest [AjPSeq] sequence object to be updated
** @param [r] src  [AjPSeq] source sequence objcet for accuracy values
** @param [r] offset [ajint] start point in the src sequence where the
**                           dest sequence was copied from
**
** @return [void]
******************************************************************************/

void ajSeqAssignAccuracy(AjPSeq dest, AjPSeq src, ajint offset)
{

    dest->Qualsize = ajSeqGetLenUngapped(dest);
    AJCNEW0(dest->Accuracy,dest->Qualsize);
    memmove(dest->Accuracy,src->Accuracy+offset,
	    dest->Qualsize*sizeof(float));
}

Another alternative would be to make AJCNEW0 and memmove calls in
embaln.c where we call the ajSeqNewRange function.

I'm finding it difficult to decide, I would appreciate any suggestions.

Mahmut



From ajb at ebi.ac.uk  Fri Jun 25 08:03:45 2010
From: ajb at ebi.ac.uk (ajb at ebi.ac.uk)
Date: Fri, 25 Jun 2010 13:03:45 +0100 (BST)
Subject: [emboss-dev] ajSeqNewRange, sequence accuracy values
In-Reply-To: <1277462975.7755.25.camel@emboss1.ebi.ac.uk>
References: <1277462975.7755.25.camel@emboss1.ebi.ac.uk>
Message-ID: <53051.86.26.12.63.1277467425.squirrel@webmail.ebi.ac.uk>

Hi Mahmut,

As a heuristic: In cases where you find you'd need to access AJAX
 object internals from either nucleus (or an application), then that
 is a sign that a new AJAX function of some sort is required.
 [assuming there isn't a preferred way of doing something which you've
  missed of course]

On that basis alone I'd go for the function.

Alan


> Hi,
>
> Working on a problem I need to copy sequence accuracy values when a new
> sequence object is defined using part of a larger original sequence.
>
> Since ajSeqNewRange functions in ajseq.c doesn't have sequence object as
> input arguments (but the sequence string) I either need to define a new
> ajSeqNewRange function with the source sequence object as one of the
> inputs or I should define a new function similar to the following that
> can be called after calling the ajSeqNewRange function.
>
> /* @func ajSeqAssignAccuracy
> ************************************************
> **
> ** Copies accuracy values from src sequence to dest sequence.
> ** Assumes dest sequence was a subset of the src sequence.
> **
> ** @param [r] dest [AjPSeq] sequence object to be updated
> ** @param [r] src  [AjPSeq] source sequence objcet for accuracy values
> ** @param [r] offset [ajint] start point in the src sequence where the
> **                           dest sequence was copied from
> **
> ** @return [void]
> ******************************************************************************/
>
> void ajSeqAssignAccuracy(AjPSeq dest, AjPSeq src, ajint offset)
> {
>
>     dest->Qualsize = ajSeqGetLenUngapped(dest);
>     AJCNEW0(dest->Accuracy,dest->Qualsize);
>     memmove(dest->Accuracy,src->Accuracy+offset,
> 	    dest->Qualsize*sizeof(float));
> }
>
> Another alternative would be to make AJCNEW0 and memmove calls in
> embaln.c where we call the ajSeqNewRange function.
>
> I'm finding it difficult to decide, I would appreciate any suggestions.
>
> Mahmut
>
>
> _______________________________________________
> emboss-dev mailing list
> emboss-dev at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/emboss-dev
>



From uludag at ebi.ac.uk  Fri Jun 25 08:26:50 2010
From: uludag at ebi.ac.uk (Mahmut Uludag)
Date: Fri, 25 Jun 2010 13:26:50 +0100
Subject: [emboss-dev] ajSeqNewRange, sequence accuracy values
In-Reply-To: <53051.86.26.12.63.1277467425.squirrel@webmail.ebi.ac.uk>
References: <1277462975.7755.25.camel@emboss1.ebi.ac.uk>
	<53051.86.26.12.63.1277467425.squirrel@webmail.ebi.ac.uk>
Message-ID: <1277468810.7755.38.camel@emboss1.ebi.ac.uk>

Hi Alan,

> As a heuristic: In cases where you find you'd need to access AJAX
>  object internals from either nucleus (or an application), then that
>  is a sign that a new AJAX function of some sort is required.
>  [assuming there isn't a preferred way of doing something which you've
>   missed of course]
> 
> On that basis alone I'd go for the function.

Thanks. Did you mean the new function I copied or a new ajSeqNewRange
function with its first argument being a sequence object? If you meant
the new ajSeqAssignAccuracy function then does its name sounds correct? 

I have noticed that I should have the following conditions in the
ajSeqAssignAccuracy function rather than checking them before calling
the function.

  (src->Accuracy!=NULL && src->Qualsize>0)

Mahmut


> > Working on a problem I need to copy sequence accuracy values when a new
> > sequence object is defined using part of a larger original sequence.
> >
> > Since ajSeqNewRange functions in ajseq.c doesn't have sequence object as
> > input arguments (but the sequence string) I either need to define a new
> > ajSeqNewRange function with the source sequence object as one of the
> > inputs or I should define a new function similar to the following that
> > can be called after calling the ajSeqNewRange function.
> >
> > /* @func ajSeqAssignAccuracy
> > ************************************************
> > **
> > ** Copies accuracy values from src sequence to dest sequence.
> > ** Assumes dest sequence was a subset of the src sequence.
> > **
> > ** @param [r] dest [AjPSeq] sequence object to be updated
> > ** @param [r] src  [AjPSeq] source sequence objcet for accuracy values
> > ** @param [r] offset [ajint] start point in the src sequence where the
> > **                           dest sequence was copied from
> > **
> > ** @return [void]
> > ******************************************************************************/
> >
> > void ajSeqAssignAccuracy(AjPSeq dest, AjPSeq src, ajint offset)
> > {
> >
> >     dest->Qualsize = ajSeqGetLenUngapped(dest);
> >     AJCNEW0(dest->Accuracy,dest->Qualsize);
> >     memmove(dest->Accuracy,src->Accuracy+offset,
> > 	    dest->Qualsize*sizeof(float));
> > }
> >
> > Another alternative would be to make AJCNEW0 and memmove calls in
> > embaln.c where we call the ajSeqNewRange function.
> >
> > I'm finding it difficult to decide, I would appreciate any suggestions.
> >
> > Mahmut



From biopython at maubp.freeserve.co.uk  Thu Jun 17 17:58:03 2010
From: biopython at maubp.freeserve.co.uk (Peter)
Date: Thu, 17 Jun 2010 18:58:03 +0100
Subject: [emboss-dev] Quality scores in union and splitter output (was:
	partial GenBank)
Message-ID: <AANLkTilOfcdvpoQQdmXF5YPeNWRXTxSnpfiEHO3WKnZu@mail.gmail.com>

Hi Peter R et al,

This email was prompted by a discussion on the main EMBOSS list:

On Thu, Jun 17, 2010 at 9:38 AM, Peter Rice <pmr at ebi.ac.uk> wrote:
>
> On 16/06/2010 13:55, Roi Brodo wrote:
>>
>> After some more reading I think I can do it using union. The problem is
>> that after I create the list (of the two ranges) using yank, union dies on
>> "union terminated: Bad value for '-sequence' and no prompt". Why is that?
>> Shouldn't i use a yank file?
>
> Yes, yank and union is the correct approach.
>
> The output of yank is a list file, so the input to union should be @filename
> to read a list of sequence addresses from the file.
>
> If you just give the filename it assumes it is sequences (perhaps a fasta
> file of sequences to be joined).
>
> We will add this to our feature requests - it should be possible to make
> seqret handle ranges from circular sequences. This will be after the next
> release as it requires rewriting the way several library functions work to
> allow the circular range.

This is an interesting example and I couldn't resist working out how I
would solve it with Biopython - something like this if anyone cares:

from Bio import SeeqIO
old = SeqIO.read("input.gbk", "gb")
new = old[800000:] + old[:100000]
SeqIO.write(new, "output.gbk", "gb")

We already have several unit tests for Biopython which check some
functionality against an EMBOSS tool. I should probably try using
EMBOSS yank and union to verify our record slicing and addition...
As part of this it occurred to me to try union on FASTQ files and I
found it does not handle the quality scores properly:

$ more example.fastq
@EAS54_6_R1_2_1_413_324
CCCTTCTTGTCTTCAGCGTTTCTCC
+
;;3;;;;;;;;;;;;7;;;;;;;88
@EAS54_6_R1_2_1_540_792
TTGGCAGGCCAAGGCCGATGGATCA
+
;;;;;;;;;;;7;;;;;-;;;3;83
@EAS54_6_R1_2_1_443_348
GTTGCTTCTGGCGTGGGTGGGGGGG
+
;;;;;;;;;;;9;7;;.7;393333


$ union --version
EMBOSS:6.2.0

$ union -sequence example.fastq -sformat fastq-sanger -osformat
fastq-sanger -stdout -auto
@EAS54_6_R1_2_1_413_324
CCCTTCTTGTCTTCAGCGTTTCTCCTTGGCAGGCCAAGGCCGATGGATCAGTTGCTTCTGGCGTGGGTGGGGGGG
+
;;3;;;;;;;;;;;;7;;;;;;;88!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

The output is well formed, and has the correct quality scores
from the first entry, but has failed to include the quality scores
of the subsequence sequences (defaulting to PHRED zero,
the exclamation mark).

There is a similar issue with splitter which seems to default
to PHRED one, the double quote, for all entries:

$ splitter -size 5 -sequence example.fastq -sformat fastq-sanger
-osformat fastq-sanger -stdout -auto
@EAS54_6_R1_2_1_413_324_1-5
CCCTT
+
"""""
@EAS54_6_R1_2_1_413_324_6-10
CTTGT
+
"""""
...
[cut]

Now admittedly neither of these operations seem to be very
natural for short read data - although it might make sense
to take a FASTQ contig and shred it using splitter for
feeding into another assembly tool?

Anyway, I thought I should report these issues.

Regards,

Peter C.


From pmr at ebi.ac.uk  Thu Jun 17 21:01:16 2010
From: pmr at ebi.ac.uk (Peter Rice)
Date: Thu, 17 Jun 2010 22:01:16 +0100
Subject: [emboss-dev] Quality scores in union and splitter output (was:
	partial GenBank)
In-Reply-To: <AANLkTilOfcdvpoQQdmXF5YPeNWRXTxSnpfiEHO3WKnZu@mail.gmail.com>
References: <AANLkTilOfcdvpoQQdmXF5YPeNWRXTxSnpfiEHO3WKnZu@mail.gmail.com>
Message-ID: <4C1A8D1C.1090901@ebi.ac.uk>

On 17/06/2010 18:58, Peter C. wrote:
> Hi Peter R et al,
>
> We already have several unit tests for Biopython which check some
> functionality against an EMBOSS tool. I should probably try using
> EMBOSS yank and union to verify our record slicing and addition...
> As part of this it occurred to me to try union on FASTQ files and I
> found it does not handle the quality scores properly:

Thanks. Just in time to fix those for the release.

regards,

Peter R.


From biopython at maubp.freeserve.co.uk  Thu Jun 17 21:23:04 2010
From: biopython at maubp.freeserve.co.uk (Peter)
Date: Thu, 17 Jun 2010 22:23:04 +0100
Subject: [emboss-dev] Quality scores in union and splitter output (was:
	partial GenBank)
In-Reply-To: <4C1A8D1C.1090901@ebi.ac.uk>
References: <AANLkTilOfcdvpoQQdmXF5YPeNWRXTxSnpfiEHO3WKnZu@mail.gmail.com>
	<4C1A8D1C.1090901@ebi.ac.uk>
Message-ID: <AANLkTinJKfekPvNtT-ryKL1B5uVucssMSFYCj1R6hwMj@mail.gmail.com>

On Thu, Jun 17, 2010 at 10:01 PM, Peter Rice <pmr at ebi.ac.uk> wrote:
>
> Thanks. Just in time to fix those for the release.
>
> regards,
>
> Peter R.

Great - that was quick work :)

Peter C.


From uludag at ebi.ac.uk  Fri Jun 25 10:49:34 2010
From: uludag at ebi.ac.uk (Mahmut Uludag)
Date: Fri, 25 Jun 2010 11:49:34 +0100
Subject: [emboss-dev] ajSeqNewRange, sequence accuracy values
Message-ID: <1277462975.7755.25.camel@emboss1.ebi.ac.uk>

Hi,

Working on a problem I need to copy sequence accuracy values when a new
sequence object is defined using part of a larger original sequence.

Since ajSeqNewRange functions in ajseq.c doesn't have sequence object as
input arguments (but the sequence string) I either need to define a new
ajSeqNewRange function with the source sequence object as one of the
inputs or I should define a new function similar to the following that
can be called after calling the ajSeqNewRange function.

/* @func ajSeqAssignAccuracy
************************************************
**
** Copies accuracy values from src sequence to dest sequence.
** Assumes dest sequence was a subset of the src sequence.
**
** @param [r] dest [AjPSeq] sequence object to be updated
** @param [r] src  [AjPSeq] source sequence objcet for accuracy values
** @param [r] offset [ajint] start point in the src sequence where the
**                           dest sequence was copied from
**
** @return [void]
******************************************************************************/

void ajSeqAssignAccuracy(AjPSeq dest, AjPSeq src, ajint offset)
{

    dest->Qualsize = ajSeqGetLenUngapped(dest);
    AJCNEW0(dest->Accuracy,dest->Qualsize);
    memmove(dest->Accuracy,src->Accuracy+offset,
	    dest->Qualsize*sizeof(float));
}

Another alternative would be to make AJCNEW0 and memmove calls in
embaln.c where we call the ajSeqNewRange function.

I'm finding it difficult to decide, I would appreciate any suggestions.

Mahmut




From ajb at ebi.ac.uk  Fri Jun 25 12:03:45 2010
From: ajb at ebi.ac.uk (ajb at ebi.ac.uk)
Date: Fri, 25 Jun 2010 13:03:45 +0100 (BST)
Subject: [emboss-dev] ajSeqNewRange, sequence accuracy values
In-Reply-To: <1277462975.7755.25.camel@emboss1.ebi.ac.uk>
References: <1277462975.7755.25.camel@emboss1.ebi.ac.uk>
Message-ID: <53051.86.26.12.63.1277467425.squirrel@webmail.ebi.ac.uk>

Hi Mahmut,

As a heuristic: In cases where you find you'd need to access AJAX
 object internals from either nucleus (or an application), then that
 is a sign that a new AJAX function of some sort is required.
 [assuming there isn't a preferred way of doing something which you've
  missed of course]

On that basis alone I'd go for the function.

Alan


> Hi,
>
> Working on a problem I need to copy sequence accuracy values when a new
> sequence object is defined using part of a larger original sequence.
>
> Since ajSeqNewRange functions in ajseq.c doesn't have sequence object as
> input arguments (but the sequence string) I either need to define a new
> ajSeqNewRange function with the source sequence object as one of the
> inputs or I should define a new function similar to the following that
> can be called after calling the ajSeqNewRange function.
>
> /* @func ajSeqAssignAccuracy
> ************************************************
> **
> ** Copies accuracy values from src sequence to dest sequence.
> ** Assumes dest sequence was a subset of the src sequence.
> **
> ** @param [r] dest [AjPSeq] sequence object to be updated
> ** @param [r] src  [AjPSeq] source sequence objcet for accuracy values
> ** @param [r] offset [ajint] start point in the src sequence where the
> **                           dest sequence was copied from
> **
> ** @return [void]
> ******************************************************************************/
>
> void ajSeqAssignAccuracy(AjPSeq dest, AjPSeq src, ajint offset)
> {
>
>     dest->Qualsize = ajSeqGetLenUngapped(dest);
>     AJCNEW0(dest->Accuracy,dest->Qualsize);
>     memmove(dest->Accuracy,src->Accuracy+offset,
> 	    dest->Qualsize*sizeof(float));
> }
>
> Another alternative would be to make AJCNEW0 and memmove calls in
> embaln.c where we call the ajSeqNewRange function.
>
> I'm finding it difficult to decide, I would appreciate any suggestions.
>
> Mahmut
>
>
> _______________________________________________
> emboss-dev mailing list
> emboss-dev at lists.open-bio.org
> http://lists.open-bio.org/mailman/listinfo/emboss-dev
>




From uludag at ebi.ac.uk  Fri Jun 25 12:26:50 2010
From: uludag at ebi.ac.uk (Mahmut Uludag)
Date: Fri, 25 Jun 2010 13:26:50 +0100
Subject: [emboss-dev] ajSeqNewRange, sequence accuracy values
In-Reply-To: <53051.86.26.12.63.1277467425.squirrel@webmail.ebi.ac.uk>
References: <1277462975.7755.25.camel@emboss1.ebi.ac.uk>
	<53051.86.26.12.63.1277467425.squirrel@webmail.ebi.ac.uk>
Message-ID: <1277468810.7755.38.camel@emboss1.ebi.ac.uk>

Hi Alan,

> As a heuristic: In cases where you find you'd need to access AJAX
>  object internals from either nucleus (or an application), then that
>  is a sign that a new AJAX function of some sort is required.
>  [assuming there isn't a preferred way of doing something which you've
>   missed of course]
> 
> On that basis alone I'd go for the function.

Thanks. Did you mean the new function I copied or a new ajSeqNewRange
function with its first argument being a sequence object? If you meant
the new ajSeqAssignAccuracy function then does its name sounds correct? 

I have noticed that I should have the following conditions in the
ajSeqAssignAccuracy function rather than checking them before calling
the function.

  (src->Accuracy!=NULL && src->Qualsize>0)

Mahmut


> > Working on a problem I need to copy sequence accuracy values when a new
> > sequence object is defined using part of a larger original sequence.
> >
> > Since ajSeqNewRange functions in ajseq.c doesn't have sequence object as
> > input arguments (but the sequence string) I either need to define a new
> > ajSeqNewRange function with the source sequence object as one of the
> > inputs or I should define a new function similar to the following that
> > can be called after calling the ajSeqNewRange function.
> >
> > /* @func ajSeqAssignAccuracy
> > ************************************************
> > **
> > ** Copies accuracy values from src sequence to dest sequence.
> > ** Assumes dest sequence was a subset of the src sequence.
> > **
> > ** @param [r] dest [AjPSeq] sequence object to be updated
> > ** @param [r] src  [AjPSeq] source sequence objcet for accuracy values
> > ** @param [r] offset [ajint] start point in the src sequence where the
> > **                           dest sequence was copied from
> > **
> > ** @return [void]
> > ******************************************************************************/
> >
> > void ajSeqAssignAccuracy(AjPSeq dest, AjPSeq src, ajint offset)
> > {
> >
> >     dest->Qualsize = ajSeqGetLenUngapped(dest);
> >     AJCNEW0(dest->Accuracy,dest->Qualsize);
> >     memmove(dest->Accuracy,src->Accuracy+offset,
> > 	    dest->Qualsize*sizeof(float));
> > }
> >
> > Another alternative would be to make AJCNEW0 and memmove calls in
> > embaln.c where we call the ajSeqNewRange function.
> >
> > I'm finding it difficult to decide, I would appreciate any suggestions.
> >
> > Mahmut