Add RefSeq data files that do exists but that we don't list.

around the messed up the anopheles Xref run where sources would be used
twice.

Also make the needed changes to the conversion script.

Also renamed these sources as they now are MULTI sources (this is just a
label, but it's better than calling them 'homo_sapiens').

Added CCDS and OTTT entries for mouse

parsing.

protein_id_predicted.

Shuffle.

'Uniprot/SWISSPROT_predicted' that are assumed to exist by
UniProtParser.

Set download=N for 'RefSeq_dna_predicted' and 'RefSeq_peptide_predicted'
(these are assumed to exist by the RefSeq parsers).

whatever file name given on the command line) and convert it into a SQL
file that can be used in place of the old "populate_metadata.sql" file
found in the "sql" subdirectory.

The output from this script should be redirected to a file that you
manually run to populate your Xref database, just as you did with
"populate_metadata.sql".  This will change in the future when the
"xref_parser.pl" is made aware of "xref_config.ini".

The file was more or less automatically created from revision 1.144 of
"sql/populate_metadata.sql" (2007-05-04 14:52:46) but has been manually
modified slightly.

The idea, for now, is to use the script "xref_config2sql.pl" to create a
SQL file from this configuration file and then to use that SQL file just
as you would have used "sql/populate_metadata.sql".  This will change in
the future when the "xref_parser.pl" is made aware of "xref_config.ini".

well as with 'LOCAL:' (which isn't).

backed out changes to deal with component seq_regions mapping to more than one assembly coord_system

Check if a particular key/value pair exists before trying to store it; warn and don't store if it's already there.

    1) For some species, the SPTREMBL file was pulled in with the
       source_id of Swiss-Prot.

    2) The RefSeq DNA and peptide sources for microbat were confused
       (the peptide files had the source_id of RefSeq_dna).

    3) Same problem, but inversed, for yeast.

    4) RefSeq files for microbat were fetched from the wrong
       FTP directory at NCBI (vertibrate_other rather than
       vertibrate_mammalian).

make it so in the generation of 'xref.txt' as well (was empty string).
Thanks to sd3 for spotting this.

Readability formatting of related code.

- determine deleted genes/transcripts by comparing old and new db, rather than reading from file lists\n- this also fixes a potential issue when deleting both whole genes and individual transcripts\n- increment gene version where individual transcripts were deleted

find the release info.  Fixed.

http://www.ebi.ac.uk/GOA/uniprot_release.html

used together with pipline code ( prevent that dbcon to pipeline db will be cleared as well)