$gbRoot/etc/genbank.conf
    contains options control the alignment and loading of
    GenBank and RefSeq data.  The file is in the same format as
    .hg.conf:
    
         name=value
         
    
         var.varname=value
    
    Variables maybe referenced in any value using the syntax:
    
         ${varname}
    
    A variable must be defined before it is referenced in the file, and are expanded immediately.
    A variable definition may reference another variable.
    
    The program gbConf can be used to print the configuration file with variables
    expanded for debugging purposes.
    
cluster.rootDir - root directory on cluster filesystems.
    cluster.paraHub - Host where parasol hub is running.
    grepIndex.genbank.default - Create grepIndex files
         for databases in the specified directory.  This value should
         correspond to the value grepIndex.genbank in the
         browser hg.conf file.  The default in this name
         leaves open the possibility of per-server overrides of this
         value, however this is not currently implemented.
    gbdb.genbank - Location of genbank gbdb
         directory.  Defaults to /gbdb/genbank.
    build.server - Build server host name.  If set,
         build scripts will generate an error if not run on this server.
    $db.  Default values are set with a
    database name of default
    $db.align.window - Size of genome alignment windows, in bases.
          The genome is partitioned in segments of no more than this sizes for alignment.
     $db.align.overlap - Number of bases of overlap in the alignment
          windows.
     $db.align.maxGap - Gaps no larger than this
          value are contained within a window rather then starting a 
          new window.  This allows gaps within introns.
     $db.align.maxWinsPerJob - Maximum number
          of small windows to pack into a job.
          Windows smaller than the specified window size are packed into
          jobs until the number of bases in the jobs is approximately
          the window size.  However a large number of small jobs can be very
          slow.  This parameters sets an absolute limit on the number of
          windows in a job.
     $db.align.unplacedChroms - 
          White-space separated list of pseudo-chromosome that contain
          unplaced sequences.  Alignments will not be allowed to span
          gaps on these chromsomes. Requires a lift file to specify the
          gaps.  The names can also be patterns similar to file name
          glob patterns that are matched against chromosome names.  The
          entire pattern must be matched. The meta characters are *
          and ?.  For example *_random.
     $db.align.minUnplacedSize - Skip
          unplaced sequences smaller than this size.
     $db.serverGenome - glob pattern or file for genome
          sequences on server. This can be NIBs or a TwoBit file.
     $db.clusterGenome - glob pattern or file for genome
          sequences on cluster. This can be NIBs or a TwoBit file. The
          resulting list of files must match was what is on the server.
     $db.ooc - path to blat ooc file on cluster, or 
          no if no blat ooc is to be used.
     $db.maxIntron - blat -maxIntron value.
     $db.lift - path to lift file for genome,
          or no if none.  Lift file is used to find gaps when partitioning the
          sequence for alignment.  It does not have to contain all chromosomes.
          Sequences named gap are treated as gaps as well as gaps in
          contig placement.  A lift file generate by gapToLift is ideal.
     $db.hapRegions - path to a PSL that contains
          alignments of haplotype pseudo-chromosomes to the reference chromosomes
          or no if none.  This is used to map
          alignments between haplotype regions for the near-best in genome alignment
          mappings.  This is not access from cluster jobs.
     $db.$srcDb.$cdnaType.$orgCat.pslCDnaFilter - Arguments
           for pslCDnaFilter for various types of alignments
           or no to skip pslCDnaFilter.
           Special handling is done for the the -polyASizes option.
           This option should be supplied without an argument.  The location
           of the generated file will be added when pslCDnaFilter
           is called.
           
          The following values are recongized in the name:
$srcDb - genbank, refseq
          $cdnaType - mrna, est
          $orgCat - native, xeno
          $db.mgcTables.$host - indicates which
          MGC tables to create when loading database on $host.
          Values are no, all, or full.
          $host is the value returned by uname -n.
     $db.mgcTables.default - indicates which
          MGC tables to create when there is no host-specific setting.
     $db.upstreamGeneTbl - If specified, use this
       table to create upstream FASTA files. It is an error if the table
       doesn't exist.  If not specified, don't create upstream
       FASTAs.
     $db.upstreamMaf - If specified, create update
       MAFs using these MAF tables and organism lists, and the genes
       in $db.upstreamGeneTbl.  The value should be
       white-space separated pairs of table name and fully qualified path
       to file with order list of organisms to include in the upstream MAF.
     $db.ccds.ncbiBuild - includes NCBI build number
       (e.g. 36.3) for CCDS table building.  Specifying this enabled auto-update
       of ccdsGene and related tables.
     $db.mgc - Should MGC tables be loaded.
          Values are yes, or no.
     $db.orfeome - Should ORFeome tables be loaded.
          Values are yes, or no.
     $db.perChromTables - Set to no
          if per-chromosome alignment tables should not be created.
     $db.$srcDb.$cdnaType.$orgCat.load -
          Should cDNAs be aligned and loaded into the database for the
          specified category. Value is yes or no.
     $db.$srcDb.$cdnaType.$orgCat.align -
          Should cDNAs be aligned for the
          specified category. Value is yes or no.  This is used if it's
          desirable to align a particular category but not create a track of it.
     $db.$srcDb.$cdnaType.$orgCat.loadDesc
          Should the description table  be loaded? Value is yes or no.
          Descriptions are normally not loaded for ESTs as they are
          large and not very useful
     $db.downloadDir - directory relative to goldenPath/
           where download files are stored.
     $db.align.prefilter - Prefilter alignments in cluster jobs.
           If set to yes or not specified do prefiltering. If set to no,
           no prefiltering is done.  This is often useful for debugging.
     $db.$srcDb.mrna.blatTargetDb -
          Should BLAT targetDb two-bit files be created for database for the
          specified srcDb. Value is yes or no.