EMBOSS: Project Meeting (Nov 8th 1999)


Attendees

Sanger Centre: Peter Rice, Ian Longden, Richard Bruskiewich
HGMP: Alan Bleasby, Gary Williams, Mark Faller, Val Curwen, Sinead O'Leary
EBI:

1. Matters Arising

None

2. General progress on release 0.0.4

MSF files now work again. They may need to be higher up in the search order.

File closing needs to check for the "getz" or any other forked process ID.

Ian is converting Will Gilbert's editor MSE, adding ACD and sequence reading and writing function calls, plus conversion from VMS to Unix. MSE will be implemented as an associated application under EMBASSY.

As part of the MSE implementation, AJAX graphics have to be separated from the rest of AJAX to make a library independent of PNG and other graphics libraries so that linking non-graphics associated packages is simple.

There are some problems running MSE on SGI machines which are still being investigated.

MSE is implemented under a new embassy directory. Use "cvs checkout -P" to get it.

Installation has been modified to add makefiles for the data directories so that they are included in "make install" processing.

EMBOSS is now available from the Sanger Centre anonymous CVS server for anyone to download (read only) the full development tree.

David Martin (EMBnet Norway) has reported problems with SuSe Linux. He is reinstalling.

Gary has modified sequence ranges to accept "@filename" syntax which reads a file of start and end positions, plus simple documentation (any non-numeric text).

Syco now includes uncommon codon usage.

Applications still need feature output implemented, probably as a separate (optional) output file as the text reports are now in production use by some groups. Peter will write some sample applications to build and modify feature tables.

Richard would like an application that could read features (e.g. GFF) and dump out selected features (e.g. CDS)

Peter has modified URL processing to use HTTP 1.0 and pass the server host address and port number, which is required by HTTP 1.1 virtual servers such as the EBI's SRS server. HTTP 1.1 could be implemented, but needs additional processing as the results come back in "chunks" instead of plain text.

"transeq" writes peptide sequences, currently named as "sequence+frame" or "sequence-frame". This has caused problems with generating file names from sequence names. The preferred solution is to use "seqname_1" to "seqname_6".

Richard has some time to work on EMBOSS again and will try the Windows NT build.

3. CCP11 Course

The course was very successful, especially Val's tutorial/practical guide.

4. Interfaces

Alan has contacted Catherine Letondal about PISE. She plans an ACD to XML converted, as PISE has its own XML to PISE converter. Some perl will be noeeded, mostly very short scripts.

Catherine Letondal will be in Hinxton on December 17th for the CCP11 software developers meeting.

Richard would like an acedb interface to EMBOSS so that ACEDB can run EMBOSS applications, possibly as an EMBOSS class within ACEDB.

5. Any Other Business

Alan has compiled and linked EMBOSS with the Birkbeck Template Library, a C++ library for protein structure analysis. A few small changes to BTL were needed. The EMBOSS/BTL build has been passed to David Moss at Birkbeck for comment.

6. Next meeting

Peter is away on Monday 15th. Next meeting Monday 22nd November, 11:00am, usual place.


Peter Rice, Informatics Division, The Sanger Centre, Wellcome Trust Genome Campus, Hinxton Hall, Cambridge, CB10 1SA, UK.