EMBOSS: Project Meeting (Mon 8th November 10)


EBI: Peter Rice, Alan Bleasby, Jon Ison, Michael Schuster
Apologies: Mahmut Uludag,

1. Minutes of the last meeting

There were no meetings for the previous 3 weeks because too many of the team were away.

Minutes of the meeting of 11th October 2010 are here.

2. Maintenance etc.

2.1 Applications

Alan is looking at eprimer3 where there is a possible incorrect definition of a directory rather than a full path top a file.

2.2 Libraries

Peter has recommitted code without the missing functions which caused problems for the mEMBOSS build.

Alan recommended building under CYGWIN to see messages from undefined functions.

2.3 Other

Peter has added EDAM.obo and DRCAT.txt to the emboss/data directory. We may also need a local copy of the sequence ontology (SO) but for now we can get by using a few internally defined SO terms.

Mahmut is supporting a Japanese Jemboss user.

3. New developments

3.1 Axis2C

Alan has built a new mEMBOSS with all Windows 32-bit modifications and Axis2C. This involved installing the NASM free assembler and "mingw" as another compiler for a related install (a more fully featured compiler than Borland).

Axis2C involved installing 9 more DLLs which were compiled because the binary distributions failed to work. It was necessary to recompile without SSL support. Adding SSL may be necessary for some pharmaceutical industry users. Alan will try to reinstall with SSL to test. The problem may be connected to the OpenSSL library naming when it is compiled from source. The build fails to load an axis DLL if SSL is included.

3.2 Data access methods

Peter has committed changes to support the new data types.

Alan noted that the Ensembl API is incomplete for support of Ensembl 58. has some pending updates which are commented in the source code.

Michael is reviewing the EFUNC and EDATA documentation errors for the Ensembl API. Most are now fixed, with 6 modules to be completed. A correction has been made for a 32-bit size issue in mEMBOSS. Windows builds use the Visual Studio Express 2010 32-bit edition.

Michael was at a human genetics meeting last week. There was interest in using a virtual machine to provide a Linux VM with APIs. Alan uses Fedora which has a virtualization option. VMs include Linux or Solaris.

Peter is considering ways to easily define dbx* indexed farm files, perhaps with a database alias list.

3.3 EDAM

Jon is working on a major revision to be released as beta_10. The current release has too many top-level terms. New higher level terms have been added and the remainder reorganised into a hierarchy under a few broader concepts.

Matus and the others in Bergen are happy. There is discussion on dropping the topic branch because some topic terms are really operations. Topics were originally defined as a simple tree for the BioCatalogue but maybe top level operations terms will meet this need.

The word used in definitions have been made stricter (identification, prediction, analysis, etc.).

The format branch has 5 top level terms now. These cover XML, HTML, part-HTML, text and datatype-specific.

3.4 Data types

Michael would like a variation data type to support users of Ensembl and UniProt.

4. Administration

4.1 Advisory Board

Peter reported on last week's SAB meeting. He will send the SAB copies of the presentations, and notes on the discussions. These need to be reviewed as soon as possible and sent to the SAB members to help in preparing their report.

4.2 Interim report

The report has been submitted to BBSRC.

4.3 Fedora

Alan noted that Fedora 14 was released on schedule. There are some issues with PMCIA devices, but this is not a problem for the EMBOSS machines.

Installation has started on Mahmut's machine for 32-bit and 64-bit versions.

5.0 Documentation and training

5.1 Books

Jon is building indexes for the 3 books. We can generate an index with Word document page numbers. We need to transform these into the numbers for the typesetter's proof hard copy pages. When the indexes have been built, we can limit them to a reasonable number of words or items and highlight them in the text.

6. User queries and answers

All done.

7. AOB


8. Date Of Next Meeting

The next EMBOSS meeting will be on Monday 22nd November. Next week there is a clash with a next-generation sequence conference.