EMBOSS: Project Meeting (Mon 8th Aug 11) |
Peter has fixed recoder and silent to search only the forward strand. A further update is needed to correctly use the sequence range and direction.
Peter will add a refseq datatype for use in reading and managing reference sequence data. Input will use a scale value to determine whether to read all sequence data for a region or to read size data where the full sequence is not required.
Peter will also add an ACD datatype to support the AjPAssem assembly data type.
Alignment inputs are currently supported as sequence sets with the align: attribute set. We do not support our alignment formats for input, partly because local alignments do not include the full sequence. Users can report alignments in FASTA and other sequence formats that are available for input.
Autoheader has been in use for the past year. Some architecture definitions are used in ajdefine.h and some compiler definitions are used in ajarch.h. The C definitions for C++ compilers are replaced by AJ_BEGIN_DECL and AJ_END_DECL macros to avoid indentation problems in Emacs.
EMBASSY packages use their own autoheader config.h files. The plplot library has its own hand-edited config.h to define the available device drivers.
Alan is checking suggested configuration changes.
Alan proposed supporting release 2.64 of autoconf for developer builds. This is recent enough to have the major new features and should be available to all serious developers.
Michael would like to to inter-convert Ensembl and EMBOSS feature tables, and to convert between coordinate systems for chromosome, supercontig and contig.
The next Ensembl release is due in September and will include some updates to variation data and support for more species.
Peter has some updates that should be in a patch release. They include the SwissProt description indexing fix, a correction to use a format value from the command line for databases with more than one possible format, the patch for parsing long GFF string values, the species patch for ensemblgenomes, and the updated tcode data file.
Mahmut noted that the "Google search" links on the old and new websites return an error.
The book XML files are now preserved in a separate CVS branch.