Plans for second half of 2003

Overview

So far this year:

We have got to a state where the SMap/FMap code is reasonably stable and the flood of bugs from its introduction seems to be reducing. We have done work on the split off FMap but this is still some way from being realised. We have also done work to support the annotators at Sanger but there are still some significant outstanding pieces of work. Finally, we have spent some time supporting Lincoln Stein and he seems to be happy for the moment with acedb.

We need now to set goals for the rest of the year which make the best use of our limited resources, especially by trying to make sure that we capitalise on work done for xace by making sure that we can easily incorporate it into the split off Fmap where appropriate.

Priorities

1. Get a prototype split off FMap working this year (we need a name for this thing....I would say "Emap" but perhaps that smacks too much of Ensemble !)
2. Continue support for Sanger Annotators, in particular humpub group (Jen) worm group and we should not forget Ian Dunhams group are avid acedb users.
3. Support of Wormbase, i.e. Lincoln

The following sections give more detail and in some places give names of people who will do the work.

Split Off FMap

1. Get all existing code checked back into cvs and see where we have got to. (Simon)
2. Finish adding DAS support to server so we can run the new Fmap from DAS servers. (Ed)
3. Investigate adding support to the acedb server to be http protocol based for a DAS only version of the acedb server. There are two parts to this:

   - assess how much work it would be to make the acedb server do this (Ed)

   - find a suitable http library (Simon)

4. Do the code to make the existing new display into a separate client. (all of us)

Human group support

1. Blixem: need to continue work to improve the way it works and its reliability. There are cases where it seems no to load the requested sequences for unknown reasons. We can improve greatly its error handling and general message outputting which would help the annotators a lot. We all agree that the code is horrendous but its important to try and improve its overall quality each time we work on it. This seems to be this groups favoured tool and we need to keep them happy. (Ed, Simon)
2. User defaults need to be completely revamped as a .acedbrc directory allowing defaults on a per database basis and using a much better interface (GTK notebook would be good). We should investigate GNOME for any stanza/ini file code it may provide for supporting defaults files. (Rob)

Worm group support

1. carry on generalisation of "db feature -> seg -> method" mapping, Simon makes the good point that the method will need to know which type (as in the sense of "gene", "homol" etc.) thingy it represents but that this can be generalised a la GMap. This has lots of feedback/synergy with the split off FMap. (Ed)
2. need to provide support for creating "genes" from the genefinder interface that are of a class other than Sequence. (Ed)
3. need some kind of superclass mechanism so that they can run queries on several classes at once (e.g. Sequence, CDS and Transcript). Simon suggested this could be via a file in wspec which could be read and used by both the Query and AQL code. At the same time we should investigate the possibility of replacing the existing subclass mechanism with something less restrictive (base it on the set of tags in an object only ??) (Simon)

General programming

1. finish the method/source code for in/excluding sets of features from GFF output. We need to nail this having invested a large amount of Robs time in it. (As a side issue need to check that the missing fields in GFF output was correctly implemented as '.' and not **UNKNOWN**) (This friday, Ed, Rob, Simon)
2. Finish off the new genetic code stuff. This is almost done. (Ed)
3. Vamp up the test system to include new tests of dna/peptide/GFF dumping. and also of smap code. Simon suggested setting up our own wormbase server so we can test stuff before handing it to Lincoln, we should see if Lincoln has a test suite but note that we don't want to end up doing Lincolns testing for him ! (Rob)