EMBOSS: Project Meeting (Mon 10th July 2006)


Attendees

EBI: Peter Rice, Jon Ison, Alan Bleasby, Mahmut Uludag,
Sanger:
Visitors:
Apologies: Shaun McGlinchey Tim Carver, Rodrigo Lopez, Lisa Mullan

1. Minutes of the last meeting

Minutes of the meeting of 27th June 2006 are here.

2. Software Development

2.1 DOM Library

Alan has completed work on the DOM library. The ajdom code will be in the release as an alpha version, but not used in any applications at this stage. First use will be for graphics output.

The library has generic entity references but has AJAX function names (so there is no clash with the native library) and in that sense is not fully compliant to the DOM standard. Some of the function names are very long.

We will need a further library to read and validate the XML (and should get XMLspy too). Alan is looking for a free library, possibly xpat, but we may need to write our own eventually.

2.2 Final developments for Release 4.0.0

Alan has updated the sourceforge requests list to include the recent addition of a ragging option to digest.

Jon has:

  1. Replaced references to obsolete application names in the ajdmx.c source file.
  2. Added a sort routine to directory list processing so the domainatrix family of programs will produce consistent output file ordering and naming across all platforms. This will avoid some issues with QA testing consistency.
  3. Updated HMMERNEW qualifier names to avoid possible clashes, including renaming the "-n" option.
  4. Mdified fuzztran, fuzzpro and fuzztran to always produce an output report (suggested by Henrikki Almusa).
  5. Modified coderet to produce 3 separate output files, two for nucleotide (mRNA and CDS) and one for the protein translations. This allows us to set the sequence type for each file, and avoids mixing protein and nucleotide sequences in a single file.
  6. Modified infoseq to provide more control over output delimiter, and cleaned up some repetitive code.
  7. Modified einverted to produce a file with the sequences of inverted repeats. The sequence name includes the start/end position in the original sequence.
  8. Converted the output of prophet to align output type, and fixed a minor documentation error.
Mike's code is done. Cleaning up will wait until after the 4.0.0 release.

Peter has:

  1. Updated the EFUNCREL and EDATAREL databases for the 4.0.0 release. The new version will appear when the EBI SRS server next updates.
  2. Completed QA tests for all EMBASSY applications apart from Vienna where examples are not so easy to find - but there are some data fiels in Ivo's thesis and others in various publications.

    All other tests passed in the latest run.

  3. Renamed the old meme application (not to be released this time, but still in the QA tests). It has been renamed omeme (it clashed with the ememe wrapper!). The ememe QA test is using the latest native meme version. Jon will add a test for an EMBOSS_MEME variable so we can explicitly test each version to catch changes.
  4. Modified default prompts to include the knowntype string.
  5. Extended help output for all programs to include default values and knowntypes for all data types, as in the acdtable output.
  6. Removed the now redundant acdtable section from the text documentation. In previous releases it was included but never well formatted. Both versions are retained in the HTML documentation for now.

Peter will update the website documentation later today.

2.3 EMBOSSWIN

Renamed emboss-ms. A copy has been given to SciTegic for testing.

Administration

3.1 Release 4.0.0

The final changes for the release have to be completed by Wednesday 12th so Alan can package everything in good time for Saturday's release.

3.2 Backups

Backup disks are installed to copy data from all the emboss machines. Peter has already found them invaluable.

4. Documentation and Training

4.1 Training

Jon has had futher contact from Bela.

5. User queries and answers

The list was reviewed. All high priority issues are done in time for the release. Everything else will wait.

6. AOB

None.

7. Date Of Next Meeting

Peter is away in 2 weeks time, and we can have a break after the release. The next meeting is on Monday 7th August.