If you want to have this newsletter mailed to you or you want to make comments/suggestions about the format/content then send an email to acedb@sanger.ac.uk.
A short newsletter this month as most work has been on the new ZMap code which will form the new split off FMap client.
Highlights for this month include: Acedb can now be compiled with either GTK version 1 or 2 now that GTK2 is available on the Alphas, ZMap can now be compiled into xace/Acedb or run as a stand alone client, several significant FMap/GFF enhancements/fixes were made to support Wormbase, plus a number of fixes to Blixem to support the Otter/lace system.
Thanks to Simon Kelley we now have a provisional/testable installed GTK2 and acedb can be compiled and run with this new version of the toolkit both on linux and the Alphas.
Acedb users can choose to build with GTK1 or 2 and either from our own distributed versions of the libraries or from their own system installed ones.
The AceConn error message routines have been revamped so that just one routine returns all error messages whether they are the simple hard coded ones or the dynamically constructed "last error" messages held within each AceConn control block.
Improvements/fixes should lead to substantial throughput gains for worm builds as non-essential data is filtered out and not dumped at all. Special HOMOL_GAP segs were created to support joining up homologies with lines (in much the same way that exons are joined up but intron lines), these mean nothing in biological terms so are filtered out which hugely reduces the size of GFF output files.
A number of improvements have been made to the way the Method object and Feature object data is handled, especially for features that have a parent-child structure such as transcripts where records must be written both for the overall feature and for its constituent parts.
Fix bug in the way we position stuff that has the Align tag set, we were not calculating the length correctly. This led to the end coordinate being truncated by the offset specified in the child start as specified following the Align tag. In this case the child contained DNA and so had its own explicit length which should have been used to correctly calculate the offset and length in the child.
This feels like a missing bit of design. We need to look for DNA and use this as an explicit length of an object, if there is no DNA we need to use whichever is the longer of the position in the parent or the maximum extent of the children.
The stand alone library has been reworked so that the ZMap window code can be compiled either into xace or into the stand alone client without code modification. This will allow testing of the interface from within xace and the existing Otter/lace system in parallel with continuing development of the stand alone client.
The stand alone client is threaded and can be compiled as a C++ executable to allow use of the more extensive C++ libraries available for handling XML datastreams.
Work is also continuing on providing acedb routines to interface to the ZMap code when compiled directly into xace.
There have been a number of general build/test improvements to provide better automation of testing in line with slimming down the effort required for daily maintenance of Acedb.
You can pick up the monthly builds from: