Keep your tracks separate from the UCSC trackdb

From genomewiki
Revision as of 07:31, 2 February 2010 by Max (talk | contribs) (New page: by Mark Diekhans on mailing list: <pre> Hi Terry, I would highly suggest keeping your track meta-data in your own source tree rather than mix them in with the mirrored ucsc tree. These c...)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigationJump to search

by Mark Diekhans on mailing list:

Hi Terry,

I would highly suggest keeping your track meta-data in your own
source tree rather than mix them in with the mirrored ucsc tree.

These can then be loaded into their own trackDb and hgFindSpec
tables (say trackDb_furey and hgFindSpec_furey).  Track
definitions can then be merged with the mirrored UCSC trackDb
at run time by specified them as a comma-separated list in the hg.conf
file:

db.trackDb=trackDb_furey,trackDb

multiple trackDb tables can be merged at run time, mine currently contains:

db.trackDb=trackDb_markdAvainBacs,trackDb_markd_sfrs1,trackDb_markd,trackDb

This allows you to safely manage your tracks outside of the UCSC
CVS tree and doesn't require you to reload your tracks with a new trackDb
table is rsynced from ucsc.

Given that, your files should be left alone when you do a `cvs update -Pd',
however I would back them up just in case.  Note that `cvs update -Pd',
should always be used for updating the ucsc

see ya in DC,
Mark