From bauer at genprofile.com Tue Aug 1 04:33:09 2000 From: bauer at genprofile.com (David Bauer) Date: Tue, 01 Aug 2000 10:33:09 +0200 Subject: HCL Exceed Message-ID: <39868B45.97805EA0@genprofile.com> Hi, I have problems with the graphical output of emboss when accessing it from WinNT via HCL Exceed X11 emulator. Here are the error messages I get from testplot: Plotting sin, cos and tan for 90 degrees X protocol error: error=8 request=1 minor=0 X protocol error: error=3 request=18 minor=0 X protocol error: error=3 request=18 minor=0 X protocol error: error=3 request=18 minor=0 X protocol error: error=3 request=2 minor=0 X protocol error: error=9 request=55 minor=0 X protocol error: error=9 request=55 minor=0 X protocol error: error=9 request=14 minor=0 Error in XCreatePixmap: BadDrawable (invalid Pixmap or Window parameter). X protocol error: error=3 request=2 minor=0 X protocol error: error=3 request=2 minor=0 X protocol error: error=3 request=12 minor=0 X protocol error: error=3 request=8 minor=0 Emboss is running on FreeBSD3.1 and there are no problems with X11 output directly under FreeBSD. Only the WinNT emulation bombs. (As workaround I can save .ps and then use ghostview, which also works.) Any idea what's wrong here ? Thanks, David. -- Dr. David Bauer GenProfile AG, Max-Delbrueck-Center, Erwin-Negelein-Haus Robert-Roessle-Str. 10, D-13125 Berlin, Germany bauer at genprofile.com, Tel:49-30-94892165, FAX:49-30-94892151 From johann at egenetics.com Tue Aug 1 05:55:47 2000 From: johann at egenetics.com (Johann Visagie) Date: Tue, 1 Aug 2000 11:55:47 +0200 Subject: HCL Exceed In-Reply-To: <39868B45.97805EA0@genprofile.com>; from bauer@genprofile.com on Tue, Aug 01, 2000 at 10:33:09AM +0200 References: <39868B45.97805EA0@genprofile.com> Message-ID: <20000801115547.A29954@fling.sanbi.ac.za> David Bauer on 2000-08-01 (Tue) at 10:33:09 +0200: > > I have problems with the graphical output of emboss when accessing it from > WinNT via HCL Exceed X11 emulator. For the record, Exceed isn't an emulator, it's a full-blown X11R6 server for the win32 architecture. :-) However... > Here are the error messages I get from testplot: > > Plotting sin, cos and tan for 90 degrees > X protocol error: error=8 request=1 minor=0 [ etc. ] > > Emboss is running on FreeBSD3.1 and there are no problems with X11 output > directly under FreeBSD. Only the WinNT emulation bombs. I get exactly the same, running on a July 14 build of FreeBSD 3.5-STABLE, displaying via a demo of Exceed 6.2 running under Win2K Prof. Displaying on various other remote (Unix-based) X11 servers gives no problems. I think I've tweaked every knob Exceed has to offer - to no avail. Baffling. > (As workaround I can save .ps and then use ghostview, which also works.) Or create a .png and use xv or similar to display. BTW, I created and submitted a FreeBSD port for EMBOSS 1.0.0, which was committed last night. Next time you cvsup your ports tree you'll see it appear under /usr/ports/biology. Pity I just missed the 3.5.1 and 4.1 releases, or a precompiled EMBOSS package would've gone out on those CDs. Oh well, next release... -- Johann From ableasby at hgmp.mrc.ac.uk Tue Aug 1 11:51:43 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Tue, 1 Aug 2000 16:51:43 +0100 (BST) Subject: HCL Exceed Message-ID: <200008011551.QAA28709@tin.hgmp.mrc.ac.uk> This is something that Peter and Ian were working on. I know that Peter can now reliably reproduce the fault on his new terminal. We have only recently been able to reliably reproduce it at HGMP (I'm told.) PLPLOT suffers from colour map/tru(e)colo(u)r problems. The PLPLOT development list has only recently become active again. We'll contact that group and have a look ourselves. Alan From ableasby at hgmp.mrc.ac.uk Wed Aug 2 05:33:02 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 2 Aug 2000 10:33:02 +0100 (BST) Subject: EMBOSS/Exceed solution Message-ID: <200008020933.KAA10553@bromine.hgmp.mrc.ac.uk> Well, it works for us anyway. Val and I have had a look on two exceed setups, one where it failed and one where it worked. As we suspected it was a trucolor problem where the failure occurred. Go to the exceed configuration. Click on Screen Definition, then set Server Visual to Pseudocolour and hopefully it'll work. Alan From bauer at genprofile.com Wed Aug 2 06:47:48 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 02 Aug 2000 12:47:48 +0200 Subject: EMBOSS/Exceed solution References: <200008020933.KAA10553@bromine.hgmp.mrc.ac.uk> Message-ID: <3987FC54.C97157FC@genprofile.com> > set Server Visual to Pseudocolour and hopefully it'll work. It works ! Thanks, David. From vcurwen at hgmp.mrc.ac.uk Wed Aug 2 06:50:53 2000 From: vcurwen at hgmp.mrc.ac.uk (Val Curwen) Date: Wed, 2 Aug 2000 11:50:53 +0100 (BST) Subject: EMBOSS/Exceed solution Message-ID: <20000802105053.3F2C8415DD@mercury.hgmp.mrc.ac.uk> >>>> set Server Visual to Pseudocolour and hopefully it'll work. >>>It works ! (Noises off: Alan and Val are cheering :-) Val Curwen Bioinformatics, MRC HGMP-Resource Centre, Hinxton, Cambridge, CB10 1SB From johann at egenetics.com Wed Aug 2 10:17:29 2000 From: johann at egenetics.com (Johann Visagie) Date: Wed, 2 Aug 2000 16:17:29 +0200 Subject: EMBOSS/Exceed solution In-Reply-To: <3987FC54.C97157FC@genprofile.com>; from bauer@genprofile.com on Wed, Aug 02, 2000 at 12:47:48PM +0200 References: <200008020933.KAA10553@bromine.hgmp.mrc.ac.uk> <3987FC54.C97157FC@genprofile.com> Message-ID: <20000802161728.L53690@fling.sanbi.ac.za> David Bauer on 2000-08-02 (Wed) at 12:47:48 +0200: > > > set Server Visual to Pseudocolour and hopefully it'll work. > > It works ! Indeed, ditto. That must be one of the few knobs offered by Exceed that I _didn't_ twiddle yesterday. :-) -- Johann From ableasby at hgmp.mrc.ac.uk Wed Aug 2 10:49:47 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 2 Aug 2000 15:49:47 +0100 (BST) Subject: EMBOSS award (FreeBSD) Message-ID: <200008021449.PAA26697@bromine.hgmp.mrc.ac.uk> At the project meeting today, amongst AOB, we decided to take up Val's idea of discretionary prizes of an EMBOSS T-shirt to anyone who has made an outstanding contribution to EMBOSS. Johann is the first winner and one will be winging its way to him shortly, if he tells us his address. Alan From johann at egenetics.com Thu Aug 3 05:45:20 2000 From: johann at egenetics.com (Johann Visagie) Date: Thu, 3 Aug 2000 11:45:20 +0200 Subject: EMBOSS award (FreeBSD) In-Reply-To: <200008021449.PAA26697@bromine.hgmp.mrc.ac.uk>; from ableasby@hgmp.mrc.ac.uk on Wed, Aug 02, 2000 at 03:49:47PM +0100 References: <200008021449.PAA26697@bromine.hgmp.mrc.ac.uk> Message-ID: <20000803114520.F586@fling.sanbi.ac.za> ableasby at hgmp.mrc.ac.uk on 2000-08-02 (Wed) at 15:49:47 +0100: > > At the project meeting today, amongst AOB, we decided to take > up Val's idea of discretionary prizes of an EMBOSS T-shirt > to anyone who has made an outstanding contribution to EMBOSS. > > Johann is the first winner and one will be winging its way to > him shortly, if he tells us his address. Ummm, gosh Alan. You embarrass me. :-) Doing the port was really comparatively little effort. Credit should go to the EMBOSS team, who have produced a mature, well structured, standards compliant and therefore very portable product. The same cannot be said for most of the software tools in bioscientific fields - believe me, I've tried porting some things recently... But, uhh, thanks. :-) 'Tis much appreciated. -- Johann PS: Address: P O Box 50852 Bloubergsands 7441 Cape Town South Africa From simon.andrews at bbsrc.ac.uk Mon Aug 7 10:48:23 2000 From: simon.andrews at bbsrc.ac.uk (simon andrews) Date: Mon, 7 Aug 2000 15:48:23 +0100 Subject: tfm in EMBOSS 1.0.0 ?? Message-ID: <486CF8F0E0BCD211A6B70000F808617884D0CD@bits-exch1.cc.bbsrc.ac.uk> A quick question . . . tfm is contained within the list of applications in the current release of EMBOSS, yet there is no sign of it following installation of v1.0.0 (it wasn't there in 0.0.4 either). I have looked back through my log files and there is no mention of any compilation having failed for tfm - but then I can't find the source for it anywhere in the distribution either! Am I missing something obvious or is it just not there?? Cheers Simon. From ableasby at hgmp.mrc.ac.uk Mon Aug 7 12:09:35 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Mon, 7 Aug 2000 17:09:35 +0100 (BST) Subject: tfm in EMBOSS 1.0.0 ?? Message-ID: <200008071609.RAA06128@tin.hgmp.mrc.ac.uk> Simon, The new version 1.0.1 is now on the ftp server. That contains tfm. Version numbers will increase as changes are made from now on. Alan From johann at egenetics.com Tue Aug 8 05:14:29 2000 From: johann at egenetics.com (Johann Visagie) Date: Tue, 8 Aug 2000 11:14:29 +0200 Subject: Extract subdirectory (Was: Re: tfm in EMBOSS 1.0.0 ??) In-Reply-To: <200008071609.RAA06128@tin.hgmp.mrc.ac.uk>; from ableasby@hgmp.mrc.ac.uk on Mon, Aug 07, 2000 at 05:09:35PM +0100 References: <200008071609.RAA06128@tin.hgmp.mrc.ac.uk> Message-ID: <20000808111429.A47916@fling.sanbi.ac.za> ableasby at hgmp.mrc.ac.uk on 2000-08-07 (Mon) at 17:09:35 +0100: > > The new version 1.0.1 is now on the ftp server. That contains tfm. > Version numbers will increase as changes are made from now on. The EMBOSS-1.0.1 tarball still extracts to a subdirectory named EMBOSS-1.0.0 - this may cause no end of confusion. :-) -- Johann From bauer at genprofile.com Tue Aug 8 05:24:27 2000 From: bauer at genprofile.com (David Bauer) Date: Tue, 08 Aug 2000 11:24:27 +0200 Subject: Extract subdirectory (Was: Re: tfm in EMBOSS 1.0.0 ??) References: <200008071609.RAA06128@tin.hgmp.mrc.ac.uk> <20000808111429.A47916@fling.sanbi.ac.za> Message-ID: <398FD1CB.BE31307B@genprofile.com> Johann Visagie wrote: > > ableasby at hgmp.mrc.ac.uk on 2000-08-07 (Mon) at 17:09:35 +0100: > > > > The new version 1.0.1 is now on the ftp server. That contains tfm. > > Version numbers will increase as changes are made from now on. > > The EMBOSS-1.0.1 tarball still extracts to a subdirectory named EMBOSS-1.0.0 > - this may cause no end of confusion. :-) And there is tfm but it does not show up with wossname nor is it working. All what I get is: tfm plotorf Displays a program's help documentation manual FATAL ERROR: The documentation for program 'plotorf' was not found. Where must I download and place the documentations so tfm can find them ? -- David From ableasby at hgmp.mrc.ac.uk Tue Aug 8 11:20:48 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Tue, 8 Aug 2000 16:20:48 +0100 (BST) Subject: Extract subdirectory (Was: Re: tfm in EMBOSS 1.0.0 ??) Message-ID: <200008081520.QAA11644@bromine.hgmp.mrc.ac.uk> The extract subdirectory and lack of doc files has been fixed in EMBOSS-1.0.2. Note that tfm is currently being modified to work if a prefix is given to configure. At the moment tfm requires a standard configuration. Alan From milius at bscl.msi.umn.edu Wed Aug 9 16:02:54 2000 From: milius at bscl.msi.umn.edu (Robert Milius) Date: Wed, 9 Aug 2000 13:02:54 -0700 Subject: databases strategies Message-ID: Hello, Please excuse a perhaps naive question. For those institutions that want download and maintain their own databases (eg Swiss-Prot, PIR, GenBank, EMBL, etc) and want to keep the duplication to a mininum (whole databases), what kind of directory structure do you use (in unix)? For example, if I want to give our users local access to NCBI's BLAST and WU's BLAST, I'll need to run formatdb and pressdb/setdb to provide different formats of the same databases. Both look at the BLASTDB environmental variable to locate the databases which means that they all have to be in the same directory. For example, I can download the files from ftp://ncbi.nlm.nih.gov/blast/db/ into /usr/local/db/blast, uncompress them and run formatdb and pressdb/setdb on them, and set the BLASTDB env to point to it. Doing this I can get BLAST to run fine. Now I want to be able to use EMBOSS on the same databases using the dbiblast utility. This program creates 4 files: acnum.hit, acnum.trg, entrynam.idx, and division.lkp. The problem is that if I run dbiblast for each db in the directory, they overwrite the ones that that were just created. I suppose I can create a separate directory for each db, and symlink the files into a common BLASTDB directory. Is there a more elegant solution I'm missing? I notice that the emboss.default appears to have a great deal of flexibility. Is it possible to have the files created by dbiblast in one folder while the data is another? I have looked at the docs in http://www.uk.embnet.org/Software/EMBOSS/Usa/databases.html but must admit it isn't all that clear to me. I tried playing with the "file:" entry, hoping to point it to different databases, but haven't had much luck with it. Any insight would be appreciated. btw, thanks to everyone who are involved in producing this wonderful package!! -bob Robert P. Milius, Ph.D. 612-626-2771 (office) Basic Sciences Computing Laboratory 612-625-4433 (fax) University of Minnesota Supercomputing Institute milius at msi.umn.edu for Digital Simulation and Advanced Computation From ableasby at hgmp.mrc.ac.uk Wed Aug 9 18:35:48 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 9 Aug 2000 23:35:48 +0100 (BST) Subject: databases strategies Message-ID: <200008092235.XAA15866@bromine.hgmp.mrc.ac.uk> Bob, Anything but naive. Here is an attempt at a concise reply, however David Martin is producing an "Administrators Guide" which will hopefully be clearer than this..... Yes, with care, it is possible to index all databases of the same type in one go and therefore have just one set of the 4 index files. All database IDs must be unique though, but that's the first rule of any database. The divisions file keeps track of which IDs belong to which database. Then, for each separate database entry in emboss.defaults you can use a combination of "files" an "exclude" to include and exclude files as appropriate. For those databases in the same directory where you can't be sure that IDs aren't shared I'd suggest keeping the indexes in separate directories. The "emboss.default" file allows specification of "dir:" for the main database files and "indexdir:" for a directory where the index files for that database are kept. HTH Alan From ableasby at hgmp.mrc.ac.uk Wed Aug 9 18:50:06 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 9 Aug 2000 23:50:06 +0100 (BST) Subject: dbiflat and dbifasta Message-ID: <200008092250.XAA16086@bromine.hgmp.mrc.ac.uk> In version 1.1.0 (now on the servers) a new database indexing application has been added (dbifasta). The fasta indexing has been removed from dbiflat to the new program. The main reason for this is that there are more fasta formats than you can shake the proverbial stick at. Whereas dbiflat used to try and guess the format, dbifasta requires you to select it. It does allow more formats to be used though. Specifically: >ID >ID ACC >db:ID >db:ID ACC >x|...[|ACC]|ID Each format can have extra information on the ID line which is treated as a description. The last listed (ncbi) format assumed the ID is in the field after the last bar (|). If there is what looks like a sensible accession number in the preceding field it will also be indexed otherwise it will be ignored. Bug fixes in 1.1.0 include dbigcg now accepting SWISS format. Also, all doc files (so far written) will be installed correctly (hopefully) and TFM has been modified to look for them in any specified install directory (or the default unpack directory). The ncbi sequence reading has been amended to cope with variable field numbers. Alan From pmr at sanger.ac.uk Thu Aug 10 04:07:03 2000 From: pmr at sanger.ac.uk (Peter Rice) Date: Thu, 10 Aug 2000 09:07:03 +0100 (BST) Subject: databases strategies In-Reply-To: (message from Robert Milius on Wed, 9 Aug 2000 13:02:54 -0700) References: Message-ID: <200008100807.JAA04294@europa.sanger.ac.uk> Robert Milius writes: >Now I want to be able to use EMBOSS on the same databases using >the dbiblast utility. This program creates 4 files: acnum.hit, acnum.trg, >entrynam.idx, and division.lkp. The problem is that if I run >dbiblast for each db in the directory, they overwrite the ones that >that were just created. > >I suppose I can create a separate directory for each db, and symlink >the files into a common BLASTDB directory. Is there a more elegant >solution I'm missing? Possibly, yes! You can build the index files in a separate directory (for each database). Run dbiblast in the index directory, and tell it the database is somewhere else. The division.lkp file should include the source database directory, and if I recall correctly the database specification in emboss.default also lets you define the database and index directories to be different. We didn't pick the index file naming - it comes from Staden and the EMBL/EBI CD-ROM indexing, and is also used by the Sanger Centre's 'efetch' utility program in acedb. Of course it could be changed - one option would be to add the database name to all the files somehow to make unique names. If you only have EMBOSS it does not matter about the compatibility. If you have Staden you would be using that for indexing and reusing the files in EMBOSS. Peter -- ---------------------------------------------------------------------- Peter Rice | Informatics Division, The Sanger Centre, E-mail: pmr at sanger.ac.uk | Wellcome Trust Genome Campus, Tel: (44) 1223 494967 | Hinxton, Cambridge, CB10 1SA, England Fax: (44) 1223 494919 | URL: http://www.sanger.ac.uk/Users/pmr/ From johann at egenetics.com Thu Aug 10 04:26:08 2000 From: johann at egenetics.com (Johann Visagie) Date: Thu, 10 Aug 2000 10:26:08 +0200 Subject: Location of emboss.default Message-ID: <20000810102608.I92457@fling.sanbi.ac.za> I'm attempting to streamline the FreeBSD port of EMBOSS a little for the next update. (The deluge of new versions is keeping me on my toes. :-) One thing I'm looking at is placing emboss.default in the more FreeBSD-ish location of /usr/local/etc. I'm currently pre-applying the following patch as part of the port installatoin process. Would this be sufficient? (Note, it shouldn't invalidate any of the other possible locations - it checks for /usr/local/etc after the others.) --- ajax/ajnam.c.orig Wed Aug 2 10:18:59 2000 +++ ajax/ajnam.c Thu Aug 10 10:06:20 2000 @@ -982,6 +982,13 @@ prefixRootFile = fopen (ajStrStr(namRootStr), "r"); } + /* look for $(PREFIX)_ROOT/../emboss.default */ + + if (!prefixRootFile) { + ajFmtPrintS (&namRootStr, "%s/etc/%s.default", prefixRoot, prefix); + prefixRootFile = fopen (ajStrStr(namRootStr), "r"); + } + if (namFileOrig) (void) ajStrAppC (&namFileOrig, ", "); (void) ajStrApp (&namFileOrig, namRootStr); -- V From jldaik at online.sh.cn Thu Aug 10 03:31:09 2000 From: jldaik at online.sh.cn (jldai) Date: Thu, 10 Aug 2000 15:31:09 +0800 Subject: for help References: <200008092250.XAA16086@bromine.hgmp.mrc.ac.uk> Message-ID: <001801c0029c$f50d09c0$9342a8c0@biodoor> hello, Please help me. I use "pressdb" to format the nt(NCBI), but this tool can not deal with large file(>2.5G). jianliang Dai M.D jldaik at online.sh.cn the genetic institute, Fudan university, Shanghai,China From johann at egenetics.com Thu Aug 10 09:29:46 2000 From: johann at egenetics.com (Johann Visagie) Date: Thu, 10 Aug 2000 15:29:46 +0200 Subject: Submitted update of EMBOSS FreeBSD port Message-ID: <20000810152945.H31126@fling.sanbi.ac.za> I've just submitted an update to the biology/emboss FreeBSD port. http://www.freebsd.org/cgi/query-pr.cgi?pr=20519 This update: - updates EMBOSS from 1.0.0 to 1.1.0 - incorporates a patch to allow the emboss.default file also to be read from /usr/local/etc, in addition to the traditional locations. - installs the provided emboss.default to /usr/local/etc as emboss.default.sample - adds a post-installation message urging the user to set the necessary environment variables and to tweak emboss.default - incorporates a patch to install the EMBOSS installed documentation (new in 1.1.0) to the FreeBSD-ish /usr/local/share/doc/EMBOSS instead of the default location. (Comments?) Still do do: - Set some defaults in emboss.default.sample which are consistent with FreeBSD diectory hierarchy before intstalling it. - Install the scripts (and other bits?) to /usr/local/share/examples/EMBOSS - Create necessary directory structure under /var/db for the user who wants to host databases locally (I need to investigate & understand better what's going on myself first! :-) -- Johann From david.martin at biotek.uio.no Thu Aug 10 11:06:02 2000 From: david.martin at biotek.uio.no (David Martin) Date: Thu, 10 Aug 2000 16:06:02 +0100 Subject: Partly updated adminguide Message-ID: A partial update of the admin guide has been uploaded to the FTP server. ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.txt This has been updated to include FreeBSD specific information (thanks Johann). Information on DBIFASTA will appear tomorrow along with some extra scripts for automatic updating of EMBOSS. comments on the text should be addressed to me. regards ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From david.martin at biotek.uio.no Thu Aug 10 11:24:50 2000 From: david.martin at biotek.uio.no (David Martin) Date: Thu, 10 Aug 2000 16:24:50 +0100 Subject: Partly updated adminguide In-Reply-To: <200008101514.QAA18902@bromine.hgmp.mrc.ac.uk> Message-ID: On Thu, 10 Aug 2000 ableasby at hgmp.mrc.ac.uk wrote: > A partial update of the admin guide has been uploaded to the FTP server. > ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.txt > > No it aint :-) It has but under a different name.. try ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.ps ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From milius at bscl.msi.umn.edu Thu Aug 10 11:25:49 2000 From: milius at bscl.msi.umn.edu (Robert Milius) Date: Thu, 10 Aug 2000 08:25:49 -0700 Subject: databases strategies In-Reply-To: <200008092235.XAA15866@bromine.hgmp.mrc.ac.uk> Message-ID: Many thanks to all who responded!! I think I've got a handle on it now... especially after perusing the "Administors Guide". thanks again! bob Robert P. Milius, Ph.D. 612-626-2771 (office) Basic Sciences Computing Laboratory 612-625-4433 (fax) University of Minnesota Supercomputing Institute milius at msi.umn.edu for Digital Simulation and Advanced Computation From david.martin at biotek.uio.no Fri Aug 11 07:11:33 2000 From: david.martin at biotek.uio.no (David Martin) Date: Fri, 11 Aug 2000 12:11:33 +0100 Subject: New draft of the admin guide Message-ID: There are one or two minor issues, but the latest draft of the admin guide is probably ready for public airing. Outstanding issues: DBIFASTA information has not been verified. I still have some problems with some databases that should in theory be processed OK. SRS lookup information has not been verified. I would appreciate feedback from those who know more about such things. Maybe Peter Rice would write a more comprehensive guide to EMBOSS-SRS integration in due course as he is the leader in that area. As it is, only the very basics are given. I am still keen to get hold of more documentation references. There are a few formatting errors outstanding. Reading the LaTeX book is the next thing on my list. Apart from that, problems with the document should be directed to me and I will endeavour to fix them. If someone creates RPM or DEB packages for EMBOSS, please let me know and I will include a description of those along with the FreeBSD description. Available from ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.ps ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From david.martin at biotek.uio.no Fri Aug 11 07:15:57 2000 From: david.martin at biotek.uio.no (David Martin) Date: Fri, 11 Aug 2000 12:15:57 +0100 Subject: New draft of the admin guide In-Reply-To: Message-ID: There is also a pdf version available at ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.pdf enjoy ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From david.martin at biotek.uio.no Fri Aug 11 09:09:21 2000 From: david.martin at biotek.uio.no (David Martin) Date: Fri, 11 Aug 2000 14:09:21 +0100 Subject: A few extra goodies in the admin guide Message-ID: I have now included my script for automated updating of my EMBOSS installation in the admin guide. It copes with the version number changes and as many packages in EMBASSY as are realistically likely to get included. New version with the same name and in the same place. It also includes two things I had overlooked: comments in emboss.default and SRSFASTA. Still the same formatting problems though. ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From david.martin at biotek.uio.no Mon Aug 14 06:46:04 2000 From: david.martin at biotek.uio.no (David Martin) Date: Mon, 14 Aug 2000 11:46:04 +0100 Subject: EMBOSS installation and configuration guide Message-ID: A draft edition of the EMBOSS administrators guide is available at ftp://ftp.no.embnet.org/pub/EMBOSS-extras/adminguide.ps or ftp://ftp.no.embnet.org/pub/EMBOSS-extras/adminguide.pdf enjoy. ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From ableasby at hgmp.mrc.ac.uk Tue Aug 15 19:01:29 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 16 Aug 2000 00:01:29 +0100 (BST) Subject: Emboss 1.2.0 now available Message-ID: <200008152301.AAA09007@tin.hgmp.mrc.ac.uk> Here are the changes in 1.2.0 Note the need to reindex blast databases with dbiblast if you use those indexes. 1. Indexing and format changes As a result of much feedback, mainly people saying this or that database can't be indexed, version 1.2.0 now handles the NCBI format cleanly. a) All library and applications using NCBI format adhere to the specifications in ftp://ncbi.nlm.nih.gov/blast/db/README N.B. This means that DBIBLAST uses NCBI format only. You will have to re-dbiblast any existing blast databases you use in EMBOSS. Why? .... EMBOSS no longer tries to guess what format a setdb/pressdb/formatdb database used to be in. Once the process has happened it is just another NCBI format database. This has the advantage that compatibility can be maintained and means that all NCBI databases can be indexed using the given keys. The other indexing programs (dbiflat, dbigcg and dbifasta) allow you to index the more friendly formats (excuse the bias). b) dbifasta has extended options. It can now cope with >db id ... e.g. KABAT "dbid" becomes a format for the method emblcd formats. Also, its NCBI option, as above, now adheres strictly to their syntax. This allows database such as pdbseq to be indexed correctly. If you wish to index non-NCBI databases using their familiar keys then don't use dbiblast. Numerous bugs in this area (all of the failure kind) have been squashed and hopefully not too many introduced. 2. New applications in 1.2.0 a) primersearch (Val Curwen) searches DNA sequences for matches with primer pairs; it reads in primer pairs from input files, searches them against any sequence(s) specified and reports all potential amplimers. You can specify a maximum percent mismatch level; for example, 10% mismatch on a primer of length 20bp means that the program will classify a primer as matching a sequence if 18 of the 20 base pairs match. b) megamerger (Gary Williams) takes two overlapping sequences and merges them into one sequence. It could thus be regarded as the opposite of what splitter does. The sequences can be very long. The program does a match of all sequence words of size 20 (by default). It then reduces this to the minimum set of overlapping matches by sorting the matches in order of size (largest size first) and then for each such match it removes any smaller matches that overlap. The result is a set of the longest ungapped alignments between the two sequences that do not overlap with each other. If the two sequences are identical in their region of overlap then there will be one region of match and no mismatches. Rgds Alan From david.martin at biotek.uio.no Thu Aug 17 07:29:04 2000 From: david.martin at biotek.uio.no (David Martin) Date: Thu, 17 Aug 2000 12:29:04 +0100 Subject: Admin guide upgrades Message-ID: In line with Alans updates to DBIBLAST, and the imminent inclusion of one of my hacks (yes I do write code as well as the admin guide) to allow group specific EMBOSS configuration, the admin guide has been updated and some spelling and formatting errors have been rectified. the new versions are available, as ever, from ftp://ftp.no.embnet.org/pub/EMBOSS-extras/adminguide.ps and adminguide.pdf depending on the format you want. HTML may follow soon and will be available via the EMBnet Norway web site, http://www.no.embnet.org/ ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From ableasby at hgmp.mrc.ac.uk Thu Aug 17 18:18:25 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Thu, 17 Aug 2000 23:18:25 +0100 (BST) Subject: EMBOSS 1.3.0 Message-ID: <200008172218.XAA17258@tin.hgmp.mrc.ac.uk> EMBOSS 1.3.0 contains two new applications [also the EMBOSSRC environment variable directive (see the adminstrators guide) and a few minor bugfixes.] 1) Vectorstrip (Val Curwen) vectorstrip is intended to be useful for stripping vector sequence from the ends of sequences of interest. For example, if a fragment has been cloned into a vector and then sequenced, the sequence may contain vector data eg from the cloning polylinker at the 5' and 3' ends of the sequence. vectorstrip will remove these contaminating regions and output trimmed sequence ready for input into another application. vectorstrip is suitable for use with low quality sequence data as it can allow for mismatches between the sequence and the vector patterns provided. You can specify the maximum level of mismatch expected. Vector data can either be provided in a file or interactively. If presented in a file, vectorstrip will search all input sequences with all vectors listed in that file. The intention is that the user can maintain a single file for use with vectorstrip, containing all the linker sequences commonly used in the laboratory. The two patterns for each vector are searched separately against the sequence. Once the search is completed, each of the hits of the 5' sequence is paired with each of the hits of the 3' sequence and the resulting subsequences are output. For example, if the 5' sequence matches the sequence from (a) position 30-60, and(b)position 70-100, and the 3' sequence matches from 150-175, then two subsequences will be output: from 61-149, and from 101-149. The lower the quality of the sequence, the more likely multiple hits become if nonzero mismatches are accepted. Default behaviour is to report only the best matches between the vector patterns and the sequence. This means that if you specify a maximum mismatch level of 10%, but the vector patterns match the sequence with zero mismatches, the search will stop and the program will output only these "best" matches. If there are no perfect matches, the program will try searching again allowing 1 mismatch, then 2, and so on until either the patterns match the sequence or the maximum specified mismatch level is exceeded. You can tell vectorstrip to show all possible matches up to your specified maximum level. 2. Diffseq (Gary Williams) diffseq takes two overlapping, nearly identical sequences and reports the differences between them, together with any features that overlap with these regions. GFF files of the differences in each sequence are also produced. diffseq should be of value when looking for SNPs, differences between strains of an organism and anything else that requires the differences between sequences to be highlighted. The sequences can be very long. The program does a match of all sequence words of size 10 (by default). It then reduces this to the minimum set of overlapping matches by sorting the matches in order of size (largest size first) and then for each such match it removes any smaller matches that overlap. The result is a set of the longest ungapped alignments between the two sequences that do not overlap with each other. The mismatched regions between these matches are reported. Alan From david.martin at biotek.uio.no Fri Aug 18 09:44:26 2000 From: david.martin at biotek.uio.no (David Martin) Date: Fri, 18 Aug 2000 14:44:26 +0100 Subject: Adminguide HTML version. Message-ID: A very rough HTML version (straight LaTeX2HTML so it is missing the pretty images, recycled electrons available upon request for you to create your own) is available as a tarball on the ftp site or can be viewed online at http://www.no.embnet.org/Programs/SAL/EMBOSS/adminguide/index.html This is now concurrent with the PS and PDF versions and includes details of the new EMBOSSRC environment variable. enjoy. All feedback gratefully received. ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From ableasby at hgmp.mrc.ac.uk Fri Aug 18 09:56:42 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Fri, 18 Aug 2000 14:56:42 +0100 (BST) Subject: EMBOSS 1.3.1 Message-ID: <200008181356.OAA01396@tin.hgmp.mrc.ac.uk> Quite a lot of "stable" releases lately. We are getting close to a more general announcement though. 1.3.1 fixes a problem in indexing when long filenames are used. It also includes a patch from Nicolas Joly for the word matching NUCLEUS code (compilation problems) and addition of GFF code for DIFFSEQ. Alan From ableasby at hgmp.mrc.ac.uk Mon Aug 21 11:32:01 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Mon, 21 Aug 2000 16:32:01 +0100 (BST) Subject: EMBOSS 1.4.0 Message-ID: <200008211532.QAA29412@tin.hgmp.mrc.ac.uk> Addition of: SILENT: Finds silent mutations which will introduce a restriction site Modifications to: RESTRICT: There are two more optional qualifiers -fragments shows lengths of fragments generated as a sorted list -name puts the name of the sequence at the end of each line Also a minor modification to stop a compilation problem on SuSE Linux 6.4 (puts gd.h in /usr/include/gd). Alan From bauer at genprofile.com Wed Aug 23 07:55:08 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 23 Aug 2000 13:55:08 +0200 Subject: emboss gui Message-ID: <39A3BB9C.8BA6DAC8@genprofile.com> Hi, what is the current status of web and graphical user interfaces to emboss? I'm just trying to put up a Pise for the users at our institute. But the current version of Pise 5.a is not synced with emboss. It seems still to be at the 0.0.4 level with many programs missing and a lot of options not available. Also the mandatory entry of an e-mail is a nightmare. What about other www and X11 things like DisGuise? David. From ableasby at hgmp.mrc.ac.uk Wed Aug 23 08:02:14 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 23 Aug 2000 13:02:14 +0100 (BST) Subject: emboss gui Message-ID: <200008231202.NAA28301@bromine.hgmp.mrc.ac.uk> Well, PISE is one of the things to get back to asap. There's just been a million and one other things to do lately. We are in contact with Rodger Stadens group and Kathryn Beal there is producing a very swish graphical interface to EMBOSS. I've been over to LMB twice to see it. Kathryn and I have arranged to get together next week to spend a morning going through the EMBOSS programs and to try to get the outputs consistent. A lot of the next versions of EMBOSS are therefore likely to involve minor changes to output format. So, work is in progress. We've spotted several things in the XML (Kathryn uses Catherines XML) that need adjusting. Catherine has been informed and we're awaiting replies. Alan From indraneel at www.cdfd.org.in Wed Aug 23 21:24:48 2000 From: indraneel at www.cdfd.org.in (Indraneel Majumdar) Date: Wed, 23 Aug 2000 18:24:48 -0700 (PDT) Subject: emboss gui In-Reply-To: <200008231202.NAA28301@bromine.hgmp.mrc.ac.uk> Message-ID: On Wed, 23 Aug 2000 ableasby at hgmp.mrc.ac.uk wrote: > Well, PISE is one of the things to get back to asap. There's just been > a million and one other things to do lately. We are in contact with > Rodger Stadens group and Kathryn Beal there is producing a very > swish graphical interface to EMBOSS. I've been over to LMB twice to Where can I get PISE and also the new interface under development? > > Alan > I've very recently started working in Bioinformatics. Sometime back (when I had only heard about GCG and didn't know of anything called EMBOSS) I made a small program for Chaos-Game-Representation. It is an interactive program using svgalib on Linux console and can read DNA/RNA sequences, has a choice of number of points to plots (starting and ending), random plot, probability and percentage of points, shows a grid, sequences, probability for random plot, printing to PNG file etc. and is very colorful. I tested it against 33million bases and it plots that in 14 seconds and prints to file instantaneously on a normal PC. It is currently under LGPL (binary under GPL). Is it possible to include this in EMBOSS. I plan to port it to Mesa to make the program platform independent and also add a command line mode. I do not know anything about AJAX or ACD yet but am trying to learn. The program is available at http://metalab.unc.edu/pub/linux/science/biology/dnacgr-0.2.tar.gz and from my website below. Hoping that you'll like the program and include it in EMBOSS, Thanks, Indraneel /************************************************************************. # Indraneel Majumdar ? E-mail: indraneel at 123india.com # # Bioinformatics Unit (EMBNET node), ? URL: http://scorpius.iwarp.com # # Centre for DNA Fingerprinting and Diagnostics, # # Hyderabad, India - 500076 # `************************************************************************/ From njoly at medusa.sis.pasteur.fr Wed Aug 23 09:05:48 2000 From: njoly at medusa.sis.pasteur.fr (Nicolas Joly) Date: Wed, 23 Aug 2000 15:05:48 +0200 Subject: emboss gui In-Reply-To: ; from indraneel@www.cdfd.org.in on Wed, Aug 23, 2000 at 06:24:48PM -0700 References: <200008231202.NAA28301@bromine.hgmp.mrc.ac.uk> Message-ID: <20000823150548.A61360@medusa.sis.pasteur.fr> On Wed, Aug 23, 2000 at 06:24:48PM -0700, Indraneel Majumdar wrote: > On Wed, 23 Aug 2000 ableasby at hgmp.mrc.ac.uk wrote: > > > Well, PISE is one of the things to get back to asap. There's just been > > a million and one other things to do lately. We are in contact with > > Rodger Stadens group and Kathryn Beal there is producing a very > > swish graphical interface to EMBOSS. I've been over to LMB twice to > > Where can I get PISE and also the new interface under development? -- Nicolas Joly Informatique Scientifique Institut Pasteur, Paris. From bauer at genprofile.com Wed Aug 23 09:30:07 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 23 Aug 2000 15:30:07 +0200 Subject: emboss gui References: Message-ID: <39A3D1DF.88A73F7E@genprofile.com> Hi, you can find Pise at: http://www-alt.pasteur.fr/~letondal/Pise But as I mentioned it is a bit outdated. After some hacks in the xml files I got something which is working. You can have a look at it: http://ubigcg.mdh4.mdc-berlin.de:8080/ Caution, the menu is generated from wossname. I still must remove programs which are not in the current Pise version, so there are some dead links. -- David. From bauer at genprofile.com Wed Aug 23 11:09:35 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 23 Aug 2000 17:09:35 +0200 Subject: emboss gui References: Message-ID: <39A3E92F.9F7F94A4@genprofile.com> > I like that. Any chance of a tarball and installation guidelines? > > ..d I got some positive response so I will put together a tarball and some installation instructions during end of this week. David. -- Dr. David Bauer GenProfile AG, Max-Delbrueck-Center, Erwin-Negelein-Haus Robert-Roessle-Str. 10, D-13125 Berlin, Germany bauer at genprofile.com, Tel:49-30-94892165, FAX:49-30-94892151 From bauer at genprofile.com Thu Aug 24 11:23:25 2000 From: bauer at genprofile.com (David Bauer) Date: Thu, 24 Aug 2000 17:23:25 +0200 Subject: Pise Message-ID: <39A53DED.CC5931B7@genprofile.com> Hi, I have cleaned up the web site at: http://ubigcg.mdh4.mdc-berlin.de:8080 There should be no more dead links. Selection lists for graphical output are reduced to those formats which are supported by Pise. Its open for testing. Please don't believe in the 5 days lifetime of your results. If I run out of disk space, I will scratch the tmp directory without looking how old the files are :-|. This modified Pise can be downloaded via anonymous ftp from: ftp://ubigcg.mdh4.mdc-berlin.de After untaring please read the README file. It contains the original installation instructions with minor aditions. The Perl and Other directories contain all the modules needed by Pise to compile. If your web server has the traditional apache layout, you should be able to compile it without any trouble. This system was developed by Catherine Letondal from the Pasteur Institute, http://www-alt.pasteur.fr/~letondal/Pise/ David. -- Dr. David Bauer GenProfile AG, Max-Delbrueck-Center, Erwin-Negelein-Haus Robert-Roessle-Str. 10, D-13125 Berlin, Germany bauer at genprofile.com, Tel:49-30-94892165, FAX:49-30-94892151 From ableasby at hgmp.mrc.ac.uk Tue Aug 29 14:59:25 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Tue, 29 Aug 2000 19:59:25 +0100 (BST) Subject: EMBOSS 1.4.2 Message-ID: <200008291859.TAA19101@tin.hgmp.mrc.ac.uk> Two releases in one day is a record. If you already have 1.4.1 then there is little to be gained from getting 1.4.2. It, as warned, is mainly concerned with modifications to help Kathryn to code the GUI. Both contain a fix for creating Gribskov profiles (a bug introduced when we changed the datafile names). Alan From letondal at pasteur.fr Wed Aug 30 05:43:22 2000 From: letondal at pasteur.fr (Catherine Letondal) Date: Wed, 30 Aug 2000 11:43:22 +0200 Subject: emboss gui In-Reply-To: Your message of "Wed, 23 Aug 2000 13:55:08 +0200." <39A3BB9C.8BA6DAC8@genprofile.com> Message-ID: <200008300943.e7U9hMB26389@nefertiti.pasteur.fr> David Bauer writes: > I'm just trying to put up a Pise for the users at our institute. But the current > version of Pise 5.a is not synced with emboss. > It seems still to be at the 0.0.4 level with many programs missing and a lot of > options not available. Sorry, I had no time for this before my holidays. The synced version should be ready today or tomorrow. -- Catherine Letondal -- Pasteur Institute Computing Center From letondal at pasteur.fr Wed Aug 30 05:50:11 2000 From: letondal at pasteur.fr (Catherine Letondal) Date: Wed, 30 Aug 2000 11:50:11 +0200 Subject: emboss gui In-Reply-To: Your message of "Wed, 23 Aug 2000 13:55:08 +0200." <39A3BB9C.8BA6DAC8@genprofile.com> Message-ID: <200008300950.e7U9oBB28514@nefertiti.pasteur.fr> David Bauer writes: > Also the mandatory entry of an e-mail is a nightmare. It could be possible to remove e-mail entry with a NO_EMAIL variable in the lib/web.pl configuration file. This would remove the field from the HTML form. The main problem is for results delivery for long jobs: the results page would have to be bookmarked and reloaded periodically by the user. I'm going to try this out. -- Catherine Letondal -- Pasteur Institute Computing Center From letondal at pasteur.fr Wed Aug 30 08:38:19 2000 From: letondal at pasteur.fr (Catherine Letondal) Date: Wed, 30 Aug 2000 14:38:19 +0200 Subject: emboss gui In-Reply-To: Your message of "Wed, 30 Aug 2000 12:06:52 +0200." <39ACDCBC.8D5B357D@genprofile.com> Message-ID: <200008301238.e7UCcJB28974@nefertiti.pasteur.fr> David Bauer writes: > Catherine Letondal wrote: > >The main problem is for results delivery for long jobs: the results page would > >have to be bookmarked and reloaded periodically by the user. > > Or by the server? I'm not sure how NCBI is doing it with the Blast results. There you > get a page which is automatically periodically reloaded with increasing time > intervals. Some programs take several days to run (phylogenetic jobs)... and there is no simple way of indicating how long the job will last to limit a server reload to a certain time. Another solution would be to let the user optionally enter his/her email only for detached jobs : http://www-alt.pasteur.fr/~letondal/Pise/results.html > Btw. it would be nice to have a PLPLOT_LIB similar to EMBOSSDIR in Makefile.global. > I placed it in emboss.xml but this is definitely not a clean solution. Well, it depends... 1) Here, we use wrappers which define all EMBOSS variables. 2) The emboss.xml file is a proper editable configuration file. The update has been automated for convenience, but the 2 files are clearly redundant. There are other EMBOSS variables, like EMBOSS_DATA, which could be defined in the Makefile.global or emboss.xml as well. 3) Isn't it possible to have these variables defined in the $HOME configuration file of the Web server? (I don't remember its name) -- Catherine Letondal -- Pasteur Institute Computing Center From bauer at genprofile.com Wed Aug 30 11:47:25 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 30 Aug 2000 17:47:25 +0200 Subject: emboss gui References: <200008301238.e7UCcJB28974@nefertiti.pasteur.fr> Message-ID: <39AD2C8D.1A9C639A@genprofile.com> Catherine Letondal wrote: > Another solution would be to let the user optionally enter his/her email only for detached > jobs : > > http://www-alt.pasteur.fr/~letondal/Pise/results.html This looks good. > but the 2 files are clearly redundant. I'm open to where the system specific things are defined. I would just prefer having one configuration file rather then 3 (Makefile.global, web.pl and emboss.xml). Or at least two, one for the global web server and Pise settings and onther one for the various applications (like EMBOSS) running under Pise. -- David. From staffan at sgibest.sto.se.pnu.com Wed Aug 30 13:07:14 2000 From: staffan at sgibest.sto.se.pnu.com (Staffan Bergh) Date: Wed, 30 Aug 2000 19:07:14 +0200 Subject: EMBOSS 1.4.2 installation Message-ID: <200008301707.TAA14128@sgibest.sto.se.pnu.com> OK, got it all installed, but here are some things that I had to hack in the configuration/compilation. Somewhat vague at points since I'm an extremely bad note-keeper, but it may help someone ... first: # uname -a IRIX64 wille 6.5 07151432 IP27 So this is IRIX 6.5, and I want to compile -n32 ... I have z, png, gd libraries (-n32) in /usr/global/lib32 [for historical reasons, NOT in /usr/local -- SGI had this annoying habit of putting their own stuff into /usr/local so you could never be sure that a OS upgrade wouldn't squat on your painfully installed stuff ... They've stopped doing that but it'd be a real pain unravelling everything now :-(] I'm not using gcc, but SGI's own cc. First hurdle: the configure script checks for a working C++ compiler ... unnecessarily, it seems. We have something wrong with the license or installation or something, so our c++ compiler doesn't work -> configure stops. You can comment out the check for C++ with no harm done! Comment out everything between lines 1503 and 1608, inclusive. (You'll need to do the same thing for the EMBASSY apps, too) Second hurdle: configure does something strange when trying to check the versions of the graphics libraries. It seems it has this fixed path ALT_HOME (set to /usr/local with a fallback to /usr) that it uses to set LD_FLAGS and CPP_FLAGS for the test of zlib/png/gd. Since I don't have the libraries in /usr/local/lib, configure fails to find zlib et al. So you need to check ALT_HOME and possibly the other two. This is lines 4644 and 4662-3. (this also affects the TOPO configure) Third hurdle: libtool seems to drop the '-n32' passed in via setenv CFLAGS '-n32'. Easy fix, setenv CC 'cc -n32' will do it. Fourth hurdle: libtool does some fancy stuff with version strings, calculating version variables for ld. However, there is at least one library that breaks this - libplplot.so.5.99 will make libtool produce an extremely long string that causes ld32 to dump core with a bus error [this is probably a bug in SGI's linker ...] The work-round for this is tricky: 1. run make until ld32 coredumps 2. edit libtool: look for case "$version_type" in none) ;; irix) and edit the while loop test (a few lines down) to while test $loop != 90; do 3. run make again and watch the output until it gets to mv -f .libs/hsregsub.lo hsregsub.o ln -s hsregsub.o hsregsub.lo /bin/sh ../libtool --mode=link cc -n32 \ -O -L/usr/global/lib32 -o libajax.la -rpath /usr/local/lib \ -version-info 1:0:1 ajacd.lo [etc] libtool will hang here, never testing true on the while loop! So control-C out of the compile, edit libtool back to having 0 in the loop test. 4. run make again. Now it goes to the end! There must be a simpler and better way of fixing this, but my shell programming skills are too feeble :-) Having said that, it installed relatively painlessly! Less pain than compiling blast, at least ... So now we start weaning users from GCG ... Next on the agenda is a web interface! Cheers /staffan ----------------------------------------------------------------------- Staffan Bergh Pharmacia, P14:5, SE-112 87 Stockholm, Sweden email: Staffan.Bergh at eu.pnu.com + Don't let that horse eat that violin phone: (int+46) 08 695 9884 + cried Chagall's mother fax: (int+46) 08 695 4084 + but he kept right on painting mobile: (int+46) 070 698 8829 + -- Lawrence Ferlinghetti From bauer at genprofile.com Tue Aug 1 08:33:09 2000 From: bauer at genprofile.com (David Bauer) Date: Tue, 01 Aug 2000 10:33:09 +0200 Subject: HCL Exceed Message-ID: <39868B45.97805EA0@genprofile.com> Hi, I have problems with the graphical output of emboss when accessing it from WinNT via HCL Exceed X11 emulator. Here are the error messages I get from testplot: Plotting sin, cos and tan for 90 degrees X protocol error: error=8 request=1 minor=0 X protocol error: error=3 request=18 minor=0 X protocol error: error=3 request=18 minor=0 X protocol error: error=3 request=18 minor=0 X protocol error: error=3 request=2 minor=0 X protocol error: error=9 request=55 minor=0 X protocol error: error=9 request=55 minor=0 X protocol error: error=9 request=14 minor=0 Error in XCreatePixmap: BadDrawable (invalid Pixmap or Window parameter). X protocol error: error=3 request=2 minor=0 X protocol error: error=3 request=2 minor=0 X protocol error: error=3 request=12 minor=0 X protocol error: error=3 request=8 minor=0 Emboss is running on FreeBSD3.1 and there are no problems with X11 output directly under FreeBSD. Only the WinNT emulation bombs. (As workaround I can save .ps and then use ghostview, which also works.) Any idea what's wrong here ? Thanks, David. -- Dr. David Bauer GenProfile AG, Max-Delbrueck-Center, Erwin-Negelein-Haus Robert-Roessle-Str. 10, D-13125 Berlin, Germany bauer at genprofile.com, Tel:49-30-94892165, FAX:49-30-94892151 From johann at egenetics.com Tue Aug 1 09:55:47 2000 From: johann at egenetics.com (Johann Visagie) Date: Tue, 1 Aug 2000 11:55:47 +0200 Subject: HCL Exceed In-Reply-To: <39868B45.97805EA0@genprofile.com>; from bauer@genprofile.com on Tue, Aug 01, 2000 at 10:33:09AM +0200 References: <39868B45.97805EA0@genprofile.com> Message-ID: <20000801115547.A29954@fling.sanbi.ac.za> David Bauer on 2000-08-01 (Tue) at 10:33:09 +0200: > > I have problems with the graphical output of emboss when accessing it from > WinNT via HCL Exceed X11 emulator. For the record, Exceed isn't an emulator, it's a full-blown X11R6 server for the win32 architecture. :-) However... > Here are the error messages I get from testplot: > > Plotting sin, cos and tan for 90 degrees > X protocol error: error=8 request=1 minor=0 [ etc. ] > > Emboss is running on FreeBSD3.1 and there are no problems with X11 output > directly under FreeBSD. Only the WinNT emulation bombs. I get exactly the same, running on a July 14 build of FreeBSD 3.5-STABLE, displaying via a demo of Exceed 6.2 running under Win2K Prof. Displaying on various other remote (Unix-based) X11 servers gives no problems. I think I've tweaked every knob Exceed has to offer - to no avail. Baffling. > (As workaround I can save .ps and then use ghostview, which also works.) Or create a .png and use xv or similar to display. BTW, I created and submitted a FreeBSD port for EMBOSS 1.0.0, which was committed last night. Next time you cvsup your ports tree you'll see it appear under /usr/ports/biology. Pity I just missed the 3.5.1 and 4.1 releases, or a precompiled EMBOSS package would've gone out on those CDs. Oh well, next release... -- Johann From ableasby at hgmp.mrc.ac.uk Tue Aug 1 15:51:43 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Tue, 1 Aug 2000 16:51:43 +0100 (BST) Subject: HCL Exceed Message-ID: <200008011551.QAA28709@tin.hgmp.mrc.ac.uk> This is something that Peter and Ian were working on. I know that Peter can now reliably reproduce the fault on his new terminal. We have only recently been able to reliably reproduce it at HGMP (I'm told.) PLPLOT suffers from colour map/tru(e)colo(u)r problems. The PLPLOT development list has only recently become active again. We'll contact that group and have a look ourselves. Alan From ableasby at hgmp.mrc.ac.uk Wed Aug 2 09:33:02 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 2 Aug 2000 10:33:02 +0100 (BST) Subject: EMBOSS/Exceed solution Message-ID: <200008020933.KAA10553@bromine.hgmp.mrc.ac.uk> Well, it works for us anyway. Val and I have had a look on two exceed setups, one where it failed and one where it worked. As we suspected it was a trucolor problem where the failure occurred. Go to the exceed configuration. Click on Screen Definition, then set Server Visual to Pseudocolour and hopefully it'll work. Alan From bauer at genprofile.com Wed Aug 2 10:47:48 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 02 Aug 2000 12:47:48 +0200 Subject: EMBOSS/Exceed solution References: <200008020933.KAA10553@bromine.hgmp.mrc.ac.uk> Message-ID: <3987FC54.C97157FC@genprofile.com> > set Server Visual to Pseudocolour and hopefully it'll work. It works ! Thanks, David. From vcurwen at hgmp.mrc.ac.uk Wed Aug 2 10:50:53 2000 From: vcurwen at hgmp.mrc.ac.uk (Val Curwen) Date: Wed, 2 Aug 2000 11:50:53 +0100 (BST) Subject: EMBOSS/Exceed solution Message-ID: <20000802105053.3F2C8415DD@mercury.hgmp.mrc.ac.uk> >>>> set Server Visual to Pseudocolour and hopefully it'll work. >>>It works ! (Noises off: Alan and Val are cheering :-) Val Curwen Bioinformatics, MRC HGMP-Resource Centre, Hinxton, Cambridge, CB10 1SB From johann at egenetics.com Wed Aug 2 14:17:29 2000 From: johann at egenetics.com (Johann Visagie) Date: Wed, 2 Aug 2000 16:17:29 +0200 Subject: EMBOSS/Exceed solution In-Reply-To: <3987FC54.C97157FC@genprofile.com>; from bauer@genprofile.com on Wed, Aug 02, 2000 at 12:47:48PM +0200 References: <200008020933.KAA10553@bromine.hgmp.mrc.ac.uk> <3987FC54.C97157FC@genprofile.com> Message-ID: <20000802161728.L53690@fling.sanbi.ac.za> David Bauer on 2000-08-02 (Wed) at 12:47:48 +0200: > > > set Server Visual to Pseudocolour and hopefully it'll work. > > It works ! Indeed, ditto. That must be one of the few knobs offered by Exceed that I _didn't_ twiddle yesterday. :-) -- Johann From ableasby at hgmp.mrc.ac.uk Wed Aug 2 14:49:47 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 2 Aug 2000 15:49:47 +0100 (BST) Subject: EMBOSS award (FreeBSD) Message-ID: <200008021449.PAA26697@bromine.hgmp.mrc.ac.uk> At the project meeting today, amongst AOB, we decided to take up Val's idea of discretionary prizes of an EMBOSS T-shirt to anyone who has made an outstanding contribution to EMBOSS. Johann is the first winner and one will be winging its way to him shortly, if he tells us his address. Alan From johann at egenetics.com Thu Aug 3 09:45:20 2000 From: johann at egenetics.com (Johann Visagie) Date: Thu, 3 Aug 2000 11:45:20 +0200 Subject: EMBOSS award (FreeBSD) In-Reply-To: <200008021449.PAA26697@bromine.hgmp.mrc.ac.uk>; from ableasby@hgmp.mrc.ac.uk on Wed, Aug 02, 2000 at 03:49:47PM +0100 References: <200008021449.PAA26697@bromine.hgmp.mrc.ac.uk> Message-ID: <20000803114520.F586@fling.sanbi.ac.za> ableasby at hgmp.mrc.ac.uk on 2000-08-02 (Wed) at 15:49:47 +0100: > > At the project meeting today, amongst AOB, we decided to take > up Val's idea of discretionary prizes of an EMBOSS T-shirt > to anyone who has made an outstanding contribution to EMBOSS. > > Johann is the first winner and one will be winging its way to > him shortly, if he tells us his address. Ummm, gosh Alan. You embarrass me. :-) Doing the port was really comparatively little effort. Credit should go to the EMBOSS team, who have produced a mature, well structured, standards compliant and therefore very portable product. The same cannot be said for most of the software tools in bioscientific fields - believe me, I've tried porting some things recently... But, uhh, thanks. :-) 'Tis much appreciated. -- Johann PS: Address: P O Box 50852 Bloubergsands 7441 Cape Town South Africa From simon.andrews at bbsrc.ac.uk Mon Aug 7 14:48:23 2000 From: simon.andrews at bbsrc.ac.uk (simon andrews) Date: Mon, 7 Aug 2000 15:48:23 +0100 Subject: tfm in EMBOSS 1.0.0 ?? Message-ID: <486CF8F0E0BCD211A6B70000F808617884D0CD@bits-exch1.cc.bbsrc.ac.uk> A quick question . . . tfm is contained within the list of applications in the current release of EMBOSS, yet there is no sign of it following installation of v1.0.0 (it wasn't there in 0.0.4 either). I have looked back through my log files and there is no mention of any compilation having failed for tfm - but then I can't find the source for it anywhere in the distribution either! Am I missing something obvious or is it just not there?? Cheers Simon. From ableasby at hgmp.mrc.ac.uk Mon Aug 7 16:09:35 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Mon, 7 Aug 2000 17:09:35 +0100 (BST) Subject: tfm in EMBOSS 1.0.0 ?? Message-ID: <200008071609.RAA06128@tin.hgmp.mrc.ac.uk> Simon, The new version 1.0.1 is now on the ftp server. That contains tfm. Version numbers will increase as changes are made from now on. Alan From johann at egenetics.com Tue Aug 8 09:14:29 2000 From: johann at egenetics.com (Johann Visagie) Date: Tue, 8 Aug 2000 11:14:29 +0200 Subject: Extract subdirectory (Was: Re: tfm in EMBOSS 1.0.0 ??) In-Reply-To: <200008071609.RAA06128@tin.hgmp.mrc.ac.uk>; from ableasby@hgmp.mrc.ac.uk on Mon, Aug 07, 2000 at 05:09:35PM +0100 References: <200008071609.RAA06128@tin.hgmp.mrc.ac.uk> Message-ID: <20000808111429.A47916@fling.sanbi.ac.za> ableasby at hgmp.mrc.ac.uk on 2000-08-07 (Mon) at 17:09:35 +0100: > > The new version 1.0.1 is now on the ftp server. That contains tfm. > Version numbers will increase as changes are made from now on. The EMBOSS-1.0.1 tarball still extracts to a subdirectory named EMBOSS-1.0.0 - this may cause no end of confusion. :-) -- Johann From bauer at genprofile.com Tue Aug 8 09:24:27 2000 From: bauer at genprofile.com (David Bauer) Date: Tue, 08 Aug 2000 11:24:27 +0200 Subject: Extract subdirectory (Was: Re: tfm in EMBOSS 1.0.0 ??) References: <200008071609.RAA06128@tin.hgmp.mrc.ac.uk> <20000808111429.A47916@fling.sanbi.ac.za> Message-ID: <398FD1CB.BE31307B@genprofile.com> Johann Visagie wrote: > > ableasby at hgmp.mrc.ac.uk on 2000-08-07 (Mon) at 17:09:35 +0100: > > > > The new version 1.0.1 is now on the ftp server. That contains tfm. > > Version numbers will increase as changes are made from now on. > > The EMBOSS-1.0.1 tarball still extracts to a subdirectory named EMBOSS-1.0.0 > - this may cause no end of confusion. :-) And there is tfm but it does not show up with wossname nor is it working. All what I get is: tfm plotorf Displays a program's help documentation manual FATAL ERROR: The documentation for program 'plotorf' was not found. Where must I download and place the documentations so tfm can find them ? -- David From ableasby at hgmp.mrc.ac.uk Tue Aug 8 15:20:48 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Tue, 8 Aug 2000 16:20:48 +0100 (BST) Subject: Extract subdirectory (Was: Re: tfm in EMBOSS 1.0.0 ??) Message-ID: <200008081520.QAA11644@bromine.hgmp.mrc.ac.uk> The extract subdirectory and lack of doc files has been fixed in EMBOSS-1.0.2. Note that tfm is currently being modified to work if a prefix is given to configure. At the moment tfm requires a standard configuration. Alan From milius at bscl.msi.umn.edu Wed Aug 9 20:02:54 2000 From: milius at bscl.msi.umn.edu (Robert Milius) Date: Wed, 9 Aug 2000 13:02:54 -0700 Subject: databases strategies Message-ID: Hello, Please excuse a perhaps naive question. For those institutions that want download and maintain their own databases (eg Swiss-Prot, PIR, GenBank, EMBL, etc) and want to keep the duplication to a mininum (whole databases), what kind of directory structure do you use (in unix)? For example, if I want to give our users local access to NCBI's BLAST and WU's BLAST, I'll need to run formatdb and pressdb/setdb to provide different formats of the same databases. Both look at the BLASTDB environmental variable to locate the databases which means that they all have to be in the same directory. For example, I can download the files from ftp://ncbi.nlm.nih.gov/blast/db/ into /usr/local/db/blast, uncompress them and run formatdb and pressdb/setdb on them, and set the BLASTDB env to point to it. Doing this I can get BLAST to run fine. Now I want to be able to use EMBOSS on the same databases using the dbiblast utility. This program creates 4 files: acnum.hit, acnum.trg, entrynam.idx, and division.lkp. The problem is that if I run dbiblast for each db in the directory, they overwrite the ones that that were just created. I suppose I can create a separate directory for each db, and symlink the files into a common BLASTDB directory. Is there a more elegant solution I'm missing? I notice that the emboss.default appears to have a great deal of flexibility. Is it possible to have the files created by dbiblast in one folder while the data is another? I have looked at the docs in http://www.uk.embnet.org/Software/EMBOSS/Usa/databases.html but must admit it isn't all that clear to me. I tried playing with the "file:" entry, hoping to point it to different databases, but haven't had much luck with it. Any insight would be appreciated. btw, thanks to everyone who are involved in producing this wonderful package!! -bob Robert P. Milius, Ph.D. 612-626-2771 (office) Basic Sciences Computing Laboratory 612-625-4433 (fax) University of Minnesota Supercomputing Institute milius at msi.umn.edu for Digital Simulation and Advanced Computation From ableasby at hgmp.mrc.ac.uk Wed Aug 9 22:35:48 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 9 Aug 2000 23:35:48 +0100 (BST) Subject: databases strategies Message-ID: <200008092235.XAA15866@bromine.hgmp.mrc.ac.uk> Bob, Anything but naive. Here is an attempt at a concise reply, however David Martin is producing an "Administrators Guide" which will hopefully be clearer than this..... Yes, with care, it is possible to index all databases of the same type in one go and therefore have just one set of the 4 index files. All database IDs must be unique though, but that's the first rule of any database. The divisions file keeps track of which IDs belong to which database. Then, for each separate database entry in emboss.defaults you can use a combination of "files" an "exclude" to include and exclude files as appropriate. For those databases in the same directory where you can't be sure that IDs aren't shared I'd suggest keeping the indexes in separate directories. The "emboss.default" file allows specification of "dir:" for the main database files and "indexdir:" for a directory where the index files for that database are kept. HTH Alan From ableasby at hgmp.mrc.ac.uk Wed Aug 9 22:50:06 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 9 Aug 2000 23:50:06 +0100 (BST) Subject: dbiflat and dbifasta Message-ID: <200008092250.XAA16086@bromine.hgmp.mrc.ac.uk> In version 1.1.0 (now on the servers) a new database indexing application has been added (dbifasta). The fasta indexing has been removed from dbiflat to the new program. The main reason for this is that there are more fasta formats than you can shake the proverbial stick at. Whereas dbiflat used to try and guess the format, dbifasta requires you to select it. It does allow more formats to be used though. Specifically: >ID >ID ACC >db:ID >db:ID ACC >x|...[|ACC]|ID Each format can have extra information on the ID line which is treated as a description. The last listed (ncbi) format assumed the ID is in the field after the last bar (|). If there is what looks like a sensible accession number in the preceding field it will also be indexed otherwise it will be ignored. Bug fixes in 1.1.0 include dbigcg now accepting SWISS format. Also, all doc files (so far written) will be installed correctly (hopefully) and TFM has been modified to look for them in any specified install directory (or the default unpack directory). The ncbi sequence reading has been amended to cope with variable field numbers. Alan From pmr at sanger.ac.uk Thu Aug 10 08:07:03 2000 From: pmr at sanger.ac.uk (Peter Rice) Date: Thu, 10 Aug 2000 09:07:03 +0100 (BST) Subject: databases strategies In-Reply-To: (message from Robert Milius on Wed, 9 Aug 2000 13:02:54 -0700) References: Message-ID: <200008100807.JAA04294@europa.sanger.ac.uk> Robert Milius writes: >Now I want to be able to use EMBOSS on the same databases using >the dbiblast utility. This program creates 4 files: acnum.hit, acnum.trg, >entrynam.idx, and division.lkp. The problem is that if I run >dbiblast for each db in the directory, they overwrite the ones that >that were just created. > >I suppose I can create a separate directory for each db, and symlink >the files into a common BLASTDB directory. Is there a more elegant >solution I'm missing? Possibly, yes! You can build the index files in a separate directory (for each database). Run dbiblast in the index directory, and tell it the database is somewhere else. The division.lkp file should include the source database directory, and if I recall correctly the database specification in emboss.default also lets you define the database and index directories to be different. We didn't pick the index file naming - it comes from Staden and the EMBL/EBI CD-ROM indexing, and is also used by the Sanger Centre's 'efetch' utility program in acedb. Of course it could be changed - one option would be to add the database name to all the files somehow to make unique names. If you only have EMBOSS it does not matter about the compatibility. If you have Staden you would be using that for indexing and reusing the files in EMBOSS. Peter -- ---------------------------------------------------------------------- Peter Rice | Informatics Division, The Sanger Centre, E-mail: pmr at sanger.ac.uk | Wellcome Trust Genome Campus, Tel: (44) 1223 494967 | Hinxton, Cambridge, CB10 1SA, England Fax: (44) 1223 494919 | URL: http://www.sanger.ac.uk/Users/pmr/ From johann at egenetics.com Thu Aug 10 08:26:08 2000 From: johann at egenetics.com (Johann Visagie) Date: Thu, 10 Aug 2000 10:26:08 +0200 Subject: Location of emboss.default Message-ID: <20000810102608.I92457@fling.sanbi.ac.za> I'm attempting to streamline the FreeBSD port of EMBOSS a little for the next update. (The deluge of new versions is keeping me on my toes. :-) One thing I'm looking at is placing emboss.default in the more FreeBSD-ish location of /usr/local/etc. I'm currently pre-applying the following patch as part of the port installatoin process. Would this be sufficient? (Note, it shouldn't invalidate any of the other possible locations - it checks for /usr/local/etc after the others.) --- ajax/ajnam.c.orig Wed Aug 2 10:18:59 2000 +++ ajax/ajnam.c Thu Aug 10 10:06:20 2000 @@ -982,6 +982,13 @@ prefixRootFile = fopen (ajStrStr(namRootStr), "r"); } + /* look for $(PREFIX)_ROOT/../emboss.default */ + + if (!prefixRootFile) { + ajFmtPrintS (&namRootStr, "%s/etc/%s.default", prefixRoot, prefix); + prefixRootFile = fopen (ajStrStr(namRootStr), "r"); + } + if (namFileOrig) (void) ajStrAppC (&namFileOrig, ", "); (void) ajStrApp (&namFileOrig, namRootStr); -- V From jldaik at online.sh.cn Thu Aug 10 07:31:09 2000 From: jldaik at online.sh.cn (jldai) Date: Thu, 10 Aug 2000 15:31:09 +0800 Subject: for help References: <200008092250.XAA16086@bromine.hgmp.mrc.ac.uk> Message-ID: <001801c0029c$f50d09c0$9342a8c0@biodoor> hello, Please help me. I use "pressdb" to format the nt(NCBI), but this tool can not deal with large file(>2.5G). jianliang Dai M.D jldaik at online.sh.cn the genetic institute, Fudan university, Shanghai,China From johann at egenetics.com Thu Aug 10 13:29:46 2000 From: johann at egenetics.com (Johann Visagie) Date: Thu, 10 Aug 2000 15:29:46 +0200 Subject: Submitted update of EMBOSS FreeBSD port Message-ID: <20000810152945.H31126@fling.sanbi.ac.za> I've just submitted an update to the biology/emboss FreeBSD port. http://www.freebsd.org/cgi/query-pr.cgi?pr=20519 This update: - updates EMBOSS from 1.0.0 to 1.1.0 - incorporates a patch to allow the emboss.default file also to be read from /usr/local/etc, in addition to the traditional locations. - installs the provided emboss.default to /usr/local/etc as emboss.default.sample - adds a post-installation message urging the user to set the necessary environment variables and to tweak emboss.default - incorporates a patch to install the EMBOSS installed documentation (new in 1.1.0) to the FreeBSD-ish /usr/local/share/doc/EMBOSS instead of the default location. (Comments?) Still do do: - Set some defaults in emboss.default.sample which are consistent with FreeBSD diectory hierarchy before intstalling it. - Install the scripts (and other bits?) to /usr/local/share/examples/EMBOSS - Create necessary directory structure under /var/db for the user who wants to host databases locally (I need to investigate & understand better what's going on myself first! :-) -- Johann From david.martin at biotek.uio.no Thu Aug 10 15:06:02 2000 From: david.martin at biotek.uio.no (David Martin) Date: Thu, 10 Aug 2000 16:06:02 +0100 Subject: Partly updated adminguide Message-ID: A partial update of the admin guide has been uploaded to the FTP server. ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.txt This has been updated to include FreeBSD specific information (thanks Johann). Information on DBIFASTA will appear tomorrow along with some extra scripts for automatic updating of EMBOSS. comments on the text should be addressed to me. regards ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From david.martin at biotek.uio.no Thu Aug 10 15:24:50 2000 From: david.martin at biotek.uio.no (David Martin) Date: Thu, 10 Aug 2000 16:24:50 +0100 Subject: Partly updated adminguide In-Reply-To: <200008101514.QAA18902@bromine.hgmp.mrc.ac.uk> Message-ID: On Thu, 10 Aug 2000 ableasby at hgmp.mrc.ac.uk wrote: > A partial update of the admin guide has been uploaded to the FTP server. > ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.txt > > No it aint :-) It has but under a different name.. try ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.ps ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From milius at bscl.msi.umn.edu Thu Aug 10 15:25:49 2000 From: milius at bscl.msi.umn.edu (Robert Milius) Date: Thu, 10 Aug 2000 08:25:49 -0700 Subject: databases strategies In-Reply-To: <200008092235.XAA15866@bromine.hgmp.mrc.ac.uk> Message-ID: Many thanks to all who responded!! I think I've got a handle on it now... especially after perusing the "Administors Guide". thanks again! bob Robert P. Milius, Ph.D. 612-626-2771 (office) Basic Sciences Computing Laboratory 612-625-4433 (fax) University of Minnesota Supercomputing Institute milius at msi.umn.edu for Digital Simulation and Advanced Computation From david.martin at biotek.uio.no Fri Aug 11 11:11:33 2000 From: david.martin at biotek.uio.no (David Martin) Date: Fri, 11 Aug 2000 12:11:33 +0100 Subject: New draft of the admin guide Message-ID: There are one or two minor issues, but the latest draft of the admin guide is probably ready for public airing. Outstanding issues: DBIFASTA information has not been verified. I still have some problems with some databases that should in theory be processed OK. SRS lookup information has not been verified. I would appreciate feedback from those who know more about such things. Maybe Peter Rice would write a more comprehensive guide to EMBOSS-SRS integration in due course as he is the leader in that area. As it is, only the very basics are given. I am still keen to get hold of more documentation references. There are a few formatting errors outstanding. Reading the LaTeX book is the next thing on my list. Apart from that, problems with the document should be directed to me and I will endeavour to fix them. If someone creates RPM or DEB packages for EMBOSS, please let me know and I will include a description of those along with the FreeBSD description. Available from ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.ps ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From david.martin at biotek.uio.no Fri Aug 11 11:15:57 2000 From: david.martin at biotek.uio.no (David Martin) Date: Fri, 11 Aug 2000 12:15:57 +0100 Subject: New draft of the admin guide In-Reply-To: Message-ID: There is also a pdf version available at ftp://ftp.no.embnet.org/embnet/tutorials/adminguide.pdf enjoy ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From david.martin at biotek.uio.no Fri Aug 11 13:09:21 2000 From: david.martin at biotek.uio.no (David Martin) Date: Fri, 11 Aug 2000 14:09:21 +0100 Subject: A few extra goodies in the admin guide Message-ID: I have now included my script for automated updating of my EMBOSS installation in the admin guide. It copes with the version number changes and as many packages in EMBASSY as are realistically likely to get included. New version with the same name and in the same place. It also includes two things I had overlooked: comments in emboss.default and SRSFASTA. Still the same formatting problems though. ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From david.martin at biotek.uio.no Mon Aug 14 10:46:04 2000 From: david.martin at biotek.uio.no (David Martin) Date: Mon, 14 Aug 2000 11:46:04 +0100 Subject: EMBOSS installation and configuration guide Message-ID: A draft edition of the EMBOSS administrators guide is available at ftp://ftp.no.embnet.org/pub/EMBOSS-extras/adminguide.ps or ftp://ftp.no.embnet.org/pub/EMBOSS-extras/adminguide.pdf enjoy. ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From ableasby at hgmp.mrc.ac.uk Tue Aug 15 23:01:29 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 16 Aug 2000 00:01:29 +0100 (BST) Subject: Emboss 1.2.0 now available Message-ID: <200008152301.AAA09007@tin.hgmp.mrc.ac.uk> Here are the changes in 1.2.0 Note the need to reindex blast databases with dbiblast if you use those indexes. 1. Indexing and format changes As a result of much feedback, mainly people saying this or that database can't be indexed, version 1.2.0 now handles the NCBI format cleanly. a) All library and applications using NCBI format adhere to the specifications in ftp://ncbi.nlm.nih.gov/blast/db/README N.B. This means that DBIBLAST uses NCBI format only. You will have to re-dbiblast any existing blast databases you use in EMBOSS. Why? .... EMBOSS no longer tries to guess what format a setdb/pressdb/formatdb database used to be in. Once the process has happened it is just another NCBI format database. This has the advantage that compatibility can be maintained and means that all NCBI databases can be indexed using the given keys. The other indexing programs (dbiflat, dbigcg and dbifasta) allow you to index the more friendly formats (excuse the bias). b) dbifasta has extended options. It can now cope with >db id ... e.g. KABAT "dbid" becomes a format for the method emblcd formats. Also, its NCBI option, as above, now adheres strictly to their syntax. This allows database such as pdbseq to be indexed correctly. If you wish to index non-NCBI databases using their familiar keys then don't use dbiblast. Numerous bugs in this area (all of the failure kind) have been squashed and hopefully not too many introduced. 2. New applications in 1.2.0 a) primersearch (Val Curwen) searches DNA sequences for matches with primer pairs; it reads in primer pairs from input files, searches them against any sequence(s) specified and reports all potential amplimers. You can specify a maximum percent mismatch level; for example, 10% mismatch on a primer of length 20bp means that the program will classify a primer as matching a sequence if 18 of the 20 base pairs match. b) megamerger (Gary Williams) takes two overlapping sequences and merges them into one sequence. It could thus be regarded as the opposite of what splitter does. The sequences can be very long. The program does a match of all sequence words of size 20 (by default). It then reduces this to the minimum set of overlapping matches by sorting the matches in order of size (largest size first) and then for each such match it removes any smaller matches that overlap. The result is a set of the longest ungapped alignments between the two sequences that do not overlap with each other. If the two sequences are identical in their region of overlap then there will be one region of match and no mismatches. Rgds Alan From david.martin at biotek.uio.no Thu Aug 17 11:29:04 2000 From: david.martin at biotek.uio.no (David Martin) Date: Thu, 17 Aug 2000 12:29:04 +0100 Subject: Admin guide upgrades Message-ID: In line with Alans updates to DBIBLAST, and the imminent inclusion of one of my hacks (yes I do write code as well as the admin guide) to allow group specific EMBOSS configuration, the admin guide has been updated and some spelling and formatting errors have been rectified. the new versions are available, as ever, from ftp://ftp.no.embnet.org/pub/EMBOSS-extras/adminguide.ps and adminguide.pdf depending on the format you want. HTML may follow soon and will be available via the EMBnet Norway web site, http://www.no.embnet.org/ ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From ableasby at hgmp.mrc.ac.uk Thu Aug 17 22:18:25 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Thu, 17 Aug 2000 23:18:25 +0100 (BST) Subject: EMBOSS 1.3.0 Message-ID: <200008172218.XAA17258@tin.hgmp.mrc.ac.uk> EMBOSS 1.3.0 contains two new applications [also the EMBOSSRC environment variable directive (see the adminstrators guide) and a few minor bugfixes.] 1) Vectorstrip (Val Curwen) vectorstrip is intended to be useful for stripping vector sequence from the ends of sequences of interest. For example, if a fragment has been cloned into a vector and then sequenced, the sequence may contain vector data eg from the cloning polylinker at the 5' and 3' ends of the sequence. vectorstrip will remove these contaminating regions and output trimmed sequence ready for input into another application. vectorstrip is suitable for use with low quality sequence data as it can allow for mismatches between the sequence and the vector patterns provided. You can specify the maximum level of mismatch expected. Vector data can either be provided in a file or interactively. If presented in a file, vectorstrip will search all input sequences with all vectors listed in that file. The intention is that the user can maintain a single file for use with vectorstrip, containing all the linker sequences commonly used in the laboratory. The two patterns for each vector are searched separately against the sequence. Once the search is completed, each of the hits of the 5' sequence is paired with each of the hits of the 3' sequence and the resulting subsequences are output. For example, if the 5' sequence matches the sequence from (a) position 30-60, and(b)position 70-100, and the 3' sequence matches from 150-175, then two subsequences will be output: from 61-149, and from 101-149. The lower the quality of the sequence, the more likely multiple hits become if nonzero mismatches are accepted. Default behaviour is to report only the best matches between the vector patterns and the sequence. This means that if you specify a maximum mismatch level of 10%, but the vector patterns match the sequence with zero mismatches, the search will stop and the program will output only these "best" matches. If there are no perfect matches, the program will try searching again allowing 1 mismatch, then 2, and so on until either the patterns match the sequence or the maximum specified mismatch level is exceeded. You can tell vectorstrip to show all possible matches up to your specified maximum level. 2. Diffseq (Gary Williams) diffseq takes two overlapping, nearly identical sequences and reports the differences between them, together with any features that overlap with these regions. GFF files of the differences in each sequence are also produced. diffseq should be of value when looking for SNPs, differences between strains of an organism and anything else that requires the differences between sequences to be highlighted. The sequences can be very long. The program does a match of all sequence words of size 10 (by default). It then reduces this to the minimum set of overlapping matches by sorting the matches in order of size (largest size first) and then for each such match it removes any smaller matches that overlap. The result is a set of the longest ungapped alignments between the two sequences that do not overlap with each other. The mismatched regions between these matches are reported. Alan From david.martin at biotek.uio.no Fri Aug 18 13:44:26 2000 From: david.martin at biotek.uio.no (David Martin) Date: Fri, 18 Aug 2000 14:44:26 +0100 Subject: Adminguide HTML version. Message-ID: A very rough HTML version (straight LaTeX2HTML so it is missing the pretty images, recycled electrons available upon request for you to create your own) is available as a tarball on the ftp site or can be viewed online at http://www.no.embnet.org/Programs/SAL/EMBOSS/adminguide/index.html This is now concurrent with the PS and PDF versions and includes details of the new EMBOSSRC environment variable. enjoy. All feedback gratefully received. ..d --------------------------------------------------------------------- * Dr. David Martin Biotechnology Centre of Oslo * * Node Manager Gaustadalleen 21 * * The Norwegian EMBNet Node P.O. box 1125 Blindern * * tel +47 22 95 87 56 N-0317 Oslo * * fax +47 22 69 41 30 Norway * --------------------------------------------------------------------- From ableasby at hgmp.mrc.ac.uk Fri Aug 18 13:56:42 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Fri, 18 Aug 2000 14:56:42 +0100 (BST) Subject: EMBOSS 1.3.1 Message-ID: <200008181356.OAA01396@tin.hgmp.mrc.ac.uk> Quite a lot of "stable" releases lately. We are getting close to a more general announcement though. 1.3.1 fixes a problem in indexing when long filenames are used. It also includes a patch from Nicolas Joly for the word matching NUCLEUS code (compilation problems) and addition of GFF code for DIFFSEQ. Alan From ableasby at hgmp.mrc.ac.uk Mon Aug 21 15:32:01 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Mon, 21 Aug 2000 16:32:01 +0100 (BST) Subject: EMBOSS 1.4.0 Message-ID: <200008211532.QAA29412@tin.hgmp.mrc.ac.uk> Addition of: SILENT: Finds silent mutations which will introduce a restriction site Modifications to: RESTRICT: There are two more optional qualifiers -fragments shows lengths of fragments generated as a sorted list -name puts the name of the sequence at the end of each line Also a minor modification to stop a compilation problem on SuSE Linux 6.4 (puts gd.h in /usr/include/gd). Alan From bauer at genprofile.com Wed Aug 23 11:55:08 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 23 Aug 2000 13:55:08 +0200 Subject: emboss gui Message-ID: <39A3BB9C.8BA6DAC8@genprofile.com> Hi, what is the current status of web and graphical user interfaces to emboss? I'm just trying to put up a Pise for the users at our institute. But the current version of Pise 5.a is not synced with emboss. It seems still to be at the 0.0.4 level with many programs missing and a lot of options not available. Also the mandatory entry of an e-mail is a nightmare. What about other www and X11 things like DisGuise? David. From ableasby at hgmp.mrc.ac.uk Wed Aug 23 12:02:14 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Wed, 23 Aug 2000 13:02:14 +0100 (BST) Subject: emboss gui Message-ID: <200008231202.NAA28301@bromine.hgmp.mrc.ac.uk> Well, PISE is one of the things to get back to asap. There's just been a million and one other things to do lately. We are in contact with Rodger Stadens group and Kathryn Beal there is producing a very swish graphical interface to EMBOSS. I've been over to LMB twice to see it. Kathryn and I have arranged to get together next week to spend a morning going through the EMBOSS programs and to try to get the outputs consistent. A lot of the next versions of EMBOSS are therefore likely to involve minor changes to output format. So, work is in progress. We've spotted several things in the XML (Kathryn uses Catherines XML) that need adjusting. Catherine has been informed and we're awaiting replies. Alan From indraneel at www.cdfd.org.in Thu Aug 24 01:24:48 2000 From: indraneel at www.cdfd.org.in (Indraneel Majumdar) Date: Wed, 23 Aug 2000 18:24:48 -0700 (PDT) Subject: emboss gui In-Reply-To: <200008231202.NAA28301@bromine.hgmp.mrc.ac.uk> Message-ID: On Wed, 23 Aug 2000 ableasby at hgmp.mrc.ac.uk wrote: > Well, PISE is one of the things to get back to asap. There's just been > a million and one other things to do lately. We are in contact with > Rodger Stadens group and Kathryn Beal there is producing a very > swish graphical interface to EMBOSS. I've been over to LMB twice to Where can I get PISE and also the new interface under development? > > Alan > I've very recently started working in Bioinformatics. Sometime back (when I had only heard about GCG and didn't know of anything called EMBOSS) I made a small program for Chaos-Game-Representation. It is an interactive program using svgalib on Linux console and can read DNA/RNA sequences, has a choice of number of points to plots (starting and ending), random plot, probability and percentage of points, shows a grid, sequences, probability for random plot, printing to PNG file etc. and is very colorful. I tested it against 33million bases and it plots that in 14 seconds and prints to file instantaneously on a normal PC. It is currently under LGPL (binary under GPL). Is it possible to include this in EMBOSS. I plan to port it to Mesa to make the program platform independent and also add a command line mode. I do not know anything about AJAX or ACD yet but am trying to learn. The program is available at http://metalab.unc.edu/pub/linux/science/biology/dnacgr-0.2.tar.gz and from my website below. Hoping that you'll like the program and include it in EMBOSS, Thanks, Indraneel /************************************************************************. # Indraneel Majumdar ? E-mail: indraneel at 123india.com # # Bioinformatics Unit (EMBNET node), ? URL: http://scorpius.iwarp.com # # Centre for DNA Fingerprinting and Diagnostics, # # Hyderabad, India - 500076 # `************************************************************************/ From njoly at medusa.sis.pasteur.fr Wed Aug 23 13:05:48 2000 From: njoly at medusa.sis.pasteur.fr (Nicolas Joly) Date: Wed, 23 Aug 2000 15:05:48 +0200 Subject: emboss gui In-Reply-To: ; from indraneel@www.cdfd.org.in on Wed, Aug 23, 2000 at 06:24:48PM -0700 References: <200008231202.NAA28301@bromine.hgmp.mrc.ac.uk> Message-ID: <20000823150548.A61360@medusa.sis.pasteur.fr> On Wed, Aug 23, 2000 at 06:24:48PM -0700, Indraneel Majumdar wrote: > On Wed, 23 Aug 2000 ableasby at hgmp.mrc.ac.uk wrote: > > > Well, PISE is one of the things to get back to asap. There's just been > > a million and one other things to do lately. We are in contact with > > Rodger Stadens group and Kathryn Beal there is producing a very > > swish graphical interface to EMBOSS. I've been over to LMB twice to > > Where can I get PISE and also the new interface under development? -- Nicolas Joly Informatique Scientifique Institut Pasteur, Paris. From bauer at genprofile.com Wed Aug 23 13:30:07 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 23 Aug 2000 15:30:07 +0200 Subject: emboss gui References: Message-ID: <39A3D1DF.88A73F7E@genprofile.com> Hi, you can find Pise at: http://www-alt.pasteur.fr/~letondal/Pise But as I mentioned it is a bit outdated. After some hacks in the xml files I got something which is working. You can have a look at it: http://ubigcg.mdh4.mdc-berlin.de:8080/ Caution, the menu is generated from wossname. I still must remove programs which are not in the current Pise version, so there are some dead links. -- David. From bauer at genprofile.com Wed Aug 23 15:09:35 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 23 Aug 2000 17:09:35 +0200 Subject: emboss gui References: Message-ID: <39A3E92F.9F7F94A4@genprofile.com> > I like that. Any chance of a tarball and installation guidelines? > > ..d I got some positive response so I will put together a tarball and some installation instructions during end of this week. David. -- Dr. David Bauer GenProfile AG, Max-Delbrueck-Center, Erwin-Negelein-Haus Robert-Roessle-Str. 10, D-13125 Berlin, Germany bauer at genprofile.com, Tel:49-30-94892165, FAX:49-30-94892151 From bauer at genprofile.com Thu Aug 24 15:23:25 2000 From: bauer at genprofile.com (David Bauer) Date: Thu, 24 Aug 2000 17:23:25 +0200 Subject: Pise Message-ID: <39A53DED.CC5931B7@genprofile.com> Hi, I have cleaned up the web site at: http://ubigcg.mdh4.mdc-berlin.de:8080 There should be no more dead links. Selection lists for graphical output are reduced to those formats which are supported by Pise. Its open for testing. Please don't believe in the 5 days lifetime of your results. If I run out of disk space, I will scratch the tmp directory without looking how old the files are :-|. This modified Pise can be downloaded via anonymous ftp from: ftp://ubigcg.mdh4.mdc-berlin.de After untaring please read the README file. It contains the original installation instructions with minor aditions. The Perl and Other directories contain all the modules needed by Pise to compile. If your web server has the traditional apache layout, you should be able to compile it without any trouble. This system was developed by Catherine Letondal from the Pasteur Institute, http://www-alt.pasteur.fr/~letondal/Pise/ David. -- Dr. David Bauer GenProfile AG, Max-Delbrueck-Center, Erwin-Negelein-Haus Robert-Roessle-Str. 10, D-13125 Berlin, Germany bauer at genprofile.com, Tel:49-30-94892165, FAX:49-30-94892151 From ableasby at hgmp.mrc.ac.uk Tue Aug 29 18:59:25 2000 From: ableasby at hgmp.mrc.ac.uk (ableasby at hgmp.mrc.ac.uk) Date: Tue, 29 Aug 2000 19:59:25 +0100 (BST) Subject: EMBOSS 1.4.2 Message-ID: <200008291859.TAA19101@tin.hgmp.mrc.ac.uk> Two releases in one day is a record. If you already have 1.4.1 then there is little to be gained from getting 1.4.2. It, as warned, is mainly concerned with modifications to help Kathryn to code the GUI. Both contain a fix for creating Gribskov profiles (a bug introduced when we changed the datafile names). Alan From letondal at pasteur.fr Wed Aug 30 09:43:22 2000 From: letondal at pasteur.fr (Catherine Letondal) Date: Wed, 30 Aug 2000 11:43:22 +0200 Subject: emboss gui In-Reply-To: Your message of "Wed, 23 Aug 2000 13:55:08 +0200." <39A3BB9C.8BA6DAC8@genprofile.com> Message-ID: <200008300943.e7U9hMB26389@nefertiti.pasteur.fr> David Bauer writes: > I'm just trying to put up a Pise for the users at our institute. But the current > version of Pise 5.a is not synced with emboss. > It seems still to be at the 0.0.4 level with many programs missing and a lot of > options not available. Sorry, I had no time for this before my holidays. The synced version should be ready today or tomorrow. -- Catherine Letondal -- Pasteur Institute Computing Center From letondal at pasteur.fr Wed Aug 30 09:50:11 2000 From: letondal at pasteur.fr (Catherine Letondal) Date: Wed, 30 Aug 2000 11:50:11 +0200 Subject: emboss gui In-Reply-To: Your message of "Wed, 23 Aug 2000 13:55:08 +0200." <39A3BB9C.8BA6DAC8@genprofile.com> Message-ID: <200008300950.e7U9oBB28514@nefertiti.pasteur.fr> David Bauer writes: > Also the mandatory entry of an e-mail is a nightmare. It could be possible to remove e-mail entry with a NO_EMAIL variable in the lib/web.pl configuration file. This would remove the field from the HTML form. The main problem is for results delivery for long jobs: the results page would have to be bookmarked and reloaded periodically by the user. I'm going to try this out. -- Catherine Letondal -- Pasteur Institute Computing Center From letondal at pasteur.fr Wed Aug 30 12:38:19 2000 From: letondal at pasteur.fr (Catherine Letondal) Date: Wed, 30 Aug 2000 14:38:19 +0200 Subject: emboss gui In-Reply-To: Your message of "Wed, 30 Aug 2000 12:06:52 +0200." <39ACDCBC.8D5B357D@genprofile.com> Message-ID: <200008301238.e7UCcJB28974@nefertiti.pasteur.fr> David Bauer writes: > Catherine Letondal wrote: > >The main problem is for results delivery for long jobs: the results page would > >have to be bookmarked and reloaded periodically by the user. > > Or by the server? I'm not sure how NCBI is doing it with the Blast results. There you > get a page which is automatically periodically reloaded with increasing time > intervals. Some programs take several days to run (phylogenetic jobs)... and there is no simple way of indicating how long the job will last to limit a server reload to a certain time. Another solution would be to let the user optionally enter his/her email only for detached jobs : http://www-alt.pasteur.fr/~letondal/Pise/results.html > Btw. it would be nice to have a PLPLOT_LIB similar to EMBOSSDIR in Makefile.global. > I placed it in emboss.xml but this is definitely not a clean solution. Well, it depends... 1) Here, we use wrappers which define all EMBOSS variables. 2) The emboss.xml file is a proper editable configuration file. The update has been automated for convenience, but the 2 files are clearly redundant. There are other EMBOSS variables, like EMBOSS_DATA, which could be defined in the Makefile.global or emboss.xml as well. 3) Isn't it possible to have these variables defined in the $HOME configuration file of the Web server? (I don't remember its name) -- Catherine Letondal -- Pasteur Institute Computing Center From bauer at genprofile.com Wed Aug 30 15:47:25 2000 From: bauer at genprofile.com (David Bauer) Date: Wed, 30 Aug 2000 17:47:25 +0200 Subject: emboss gui References: <200008301238.e7UCcJB28974@nefertiti.pasteur.fr> Message-ID: <39AD2C8D.1A9C639A@genprofile.com> Catherine Letondal wrote: > Another solution would be to let the user optionally enter his/her email only for detached > jobs : > > http://www-alt.pasteur.fr/~letondal/Pise/results.html This looks good. > but the 2 files are clearly redundant. I'm open to where the system specific things are defined. I would just prefer having one configuration file rather then 3 (Makefile.global, web.pl and emboss.xml). Or at least two, one for the global web server and Pise settings and onther one for the various applications (like EMBOSS) running under Pise. -- David. From staffan at sgibest.sto.se.pnu.com Wed Aug 30 17:07:14 2000 From: staffan at sgibest.sto.se.pnu.com (Staffan Bergh) Date: Wed, 30 Aug 2000 19:07:14 +0200 Subject: EMBOSS 1.4.2 installation Message-ID: <200008301707.TAA14128@sgibest.sto.se.pnu.com> OK, got it all installed, but here are some things that I had to hack in the configuration/compilation. Somewhat vague at points since I'm an extremely bad note-keeper, but it may help someone ... first: # uname -a IRIX64 wille 6.5 07151432 IP27 So this is IRIX 6.5, and I want to compile -n32 ... I have z, png, gd libraries (-n32) in /usr/global/lib32 [for historical reasons, NOT in /usr/local -- SGI had this annoying habit of putting their own stuff into /usr/local so you could never be sure that a OS upgrade wouldn't squat on your painfully installed stuff ... They've stopped doing that but it'd be a real pain unravelling everything now :-(] I'm not using gcc, but SGI's own cc. First hurdle: the configure script checks for a working C++ compiler ... unnecessarily, it seems. We have something wrong with the license or installation or something, so our c++ compiler doesn't work -> configure stops. You can comment out the check for C++ with no harm done! Comment out everything between lines 1503 and 1608, inclusive. (You'll need to do the same thing for the EMBASSY apps, too) Second hurdle: configure does something strange when trying to check the versions of the graphics libraries. It seems it has this fixed path ALT_HOME (set to /usr/local with a fallback to /usr) that it uses to set LD_FLAGS and CPP_FLAGS for the test of zlib/png/gd. Since I don't have the libraries in /usr/local/lib, configure fails to find zlib et al. So you need to check ALT_HOME and possibly the other two. This is lines 4644 and 4662-3. (this also affects the TOPO configure) Third hurdle: libtool seems to drop the '-n32' passed in via setenv CFLAGS '-n32'. Easy fix, setenv CC 'cc -n32' will do it. Fourth hurdle: libtool does some fancy stuff with version strings, calculating version variables for ld. However, there is at least one library that breaks this - libplplot.so.5.99 will make libtool produce an extremely long string that causes ld32 to dump core with a bus error [this is probably a bug in SGI's linker ...] The work-round for this is tricky: 1. run make until ld32 coredumps 2. edit libtool: look for case "$version_type" in none) ;; irix) and edit the while loop test (a few lines down) to while test $loop != 90; do 3. run make again and watch the output until it gets to mv -f .libs/hsregsub.lo hsregsub.o ln -s hsregsub.o hsregsub.lo /bin/sh ../libtool --mode=link cc -n32 \ -O -L/usr/global/lib32 -o libajax.la -rpath /usr/local/lib \ -version-info 1:0:1 ajacd.lo [etc] libtool will hang here, never testing true on the while loop! So control-C out of the compile, edit libtool back to having 0 in the loop test. 4. run make again. Now it goes to the end! There must be a simpler and better way of fixing this, but my shell programming skills are too feeble :-) Having said that, it installed relatively painlessly! Less pain than compiling blast, at least ... So now we start weaning users from GCG ... Next on the agenda is a web interface! Cheers /staffan ----------------------------------------------------------------------- Staffan Bergh Pharmacia, P14:5, SE-112 87 Stockholm, Sweden email: Staffan.Bergh at eu.pnu.com + Don't let that horse eat that violin phone: (int+46) 08 695 9884 + cried Chagall's mother fax: (int+46) 08 695 4084 + but he kept right on painting mobile: (int+46) 070 698 8829 + -- Lawrence Ferlinghetti