EMBOSS: Project Meeting (Mon 29th March 10)


Attendees

EBI: Peter Rice, Alan Bleasby, Jon Ison, Lisa Mullan
Visitors:
Apologies: Mahmut Uludag

1. Minutes of the last meeting

Minutes of the meeting of 22nd March 2010 are here.

2. Maintenance etc.

2.1 Applications

Peter has updated tfscan to produce a report as output.

2.2 Libraries

Peter has updated ACD processing of local data files to include searching local user directories. This fixes a reported problem in pepstats.

Peter has updated the handling of missing values in ACD. Past releases used '.' as a general missing value, but this is inappropriate for directories. The missing value is now an empty string, specified as '' for a parameter. A bug in processing -qualifier='' has been fixed to correctly return an empty string.

Alan has timeouts and SIGALRM working under Windows. The function name is the same but the underlying method is different.

Alan added some casts to the new ajsys.c code for compilation on Microsoft compilers and modified the use of open and read to use Microsoft preferred variants on Windows systems.

2.3 Other

Alan has updated mEMBOSS for the EBI course this week.

3. New developments

3.1 BioMart access

Alan noted that the martseqs application reports incomplete definitions from some BioMart servers as they run an older version of BioMart and fail to return all possible fields. The solution (from discussions with Syed) is to try calling the main BioMart server which will have all the metadata. The BioMart team recommend first calling the remote server as there is a potential problem of overloading the central registry if too many requests are made there.

Alan reported that the sequence type can be inferred from the name of an attribute for current marts. We can make a list of known sequence attribute names to generate a sequence type of 'nucleotide' or 'protein'.

3. EDAM

Jon has been working with Matus Kalas (U. Bergen) on annotation of the BioXSD schema with EDAM terms. In most cases EDAM had a term that could be used. In some cases a more detailed term was added, or the BioXSD schema was modified to fit the existing terms.

The EDAM tool function and datatype branches could be simplified for use by BioXSD and by the reviewers at the Swiss Institute of Bioinformatics.

The datatypes branch should be specifically for semantic descriptions of datatypes. Terms that relate to syntax should be moved to a separate "formats" branch with relations between the formats and datatypes.

Jon mentioned possibilities for maintaining the list of databases in other formats. NO final decision was reached.

Jon has been evaluating ontology editors, especially for managing OWL definitions. One possibility is TopBraid which costs $500. Jon will check what other OBO Foundry ontologies use where they provide (or maintain) an OWL version.

Jon described the Sequence Ontology teams new browser which gives a useful graphical view of ontology terms and structure.

4. Administration

Peter noted that SourceForge CVS access went down on Thursday "for 30 minutes to perform maintenance". This was increased to "2 days" but no reports back to anyone who complained. Any bug report tickets were just flagged with "see status page" without any explanation for the 2 days (or more) loss of service.

Alan described the mount point for the internal web server. Files moved to /shared/www/html will appear in the internal web server. So far we are only using the Wiki.

A future meeting should consider requirements for the next release.

5. Documentation and Training

5.1 Books

Jon reported that the Developers Manual is now essentially complete. Remaining issues are the need to define URLs for the new book-based website, and the generation of XML library documentation from the embossdoc.pl script that creates the existing HTML library documentation.

Peter will write a script to check all URLs in the book source files.

Alan updated the Developer's Manual to cover the setting of compiler warnings and configuring SQL access through MySQL or PostGres.

Jon welcomes suggestions for the front covers. We may need to let the publishers choose something appropriate.

6. User queries and answers

All outstanding queries put on the Sourceforge tracker.

7. AOB

Lisa joined us. She is teaching EMBOSS on EBI's plant Bioinformatics course today.

8. Date Of Next Meeting

Next week is Easter. The next meeting will be on Monday 12th April.