Import of CCDS (Consensus CDS) into browser database.

Dumps of the CCDS database are in at:

     ftp://ftp-private.ncbi.nih.gov/
user ccds

named in the form:
        CCDS.20050303.tar.gz

Initially, import is done manually, later this maybe be
automated to be part of the genbank updates.


Import database:

The CCDS schema from NCBI, which I believe is Sybase, is
converted to MySQL in a semi-manual fashion.  The directory
ncbiSql contains the last NCBI sql files imported:

   createTables.sql
   createKeys.sql

After determining what has changed in the NCBI SQL files, copy them
to ncbiSql and convert them to MySQL with:

   ncbiSql/convertCreateTables ncbiSql/createTables.sql > createTables.sql
   ncbiSql/convertCreateKeys ncbiSql/createKeys.sql > createKeys.sql

result of the conversion are checked-in to CVS.   The sql is compiled
into the ccdsImport program, so it must be recompiled. These programs may
have to be modified if the schema changes.

The program ccdsImport imports the table dumps into a database ccds,
which is then used to generate the tables that are loaded into
the browser databases. For example:

    /cluster/data/genbank/bin/x86_64/ccdsImport ccds data/*.txt


The browser tables created are:
       ccdsGene - genePred track for CCDS
       ccdsInfo - cross-reference information for CCDS.

The ccdsMkTables program create the browser database tables from the imported
ccds database:

    /cluster/data/genbank/bin/x86_64/ccdsMkTables -loadDb ccds hg17 ccdsInfo ccdsGene

to load canidates:
    /cluster/data/genbank/bin/x86_64/ccdsMkTables -stat=Candidate -loadDb ccds mm8 ccdsInfo ccdsGene

  

