Keeping sandbox tracks separated on hgwdev
From genomewiki
Jump to navigationJump to search
If you want your own trackDb hierarchy whic is not in git and shows up only in your sandbox and will never appear on hgwdev there a few steps:
- add another "sandbox-only" trackDb entry to your hg.conf, after a comma, e.g. trackDb_<user>_sandbox so the whole trackDb statement reads:
db.trackDb=trackDb_<user>,trackDb_<user>_sandbox
- create a directory for your own trackDb hierarchy, e.g. ~usr/trackDb
- define the variable USER in the makefile and set it to <user>_sandbox e.g.
USER=max_sandbox
- you need to instruct loadTables to read the right sql file and modify the makefile to find the right scripts
The final makefile in ~/usr/trackDb should look like this:
USER=maxSandbox EXTRA=-sqlDir=~kent/src/hg/lib REALTRACKDBDIR=~kent/src/hg/makeDb/trackDb
update: $(REALTRACKDBDIR)/loadTracks ${EXTRA} trackDb_${USER} hgFindSpec_${USER} ${DBS} $(REALTRACKDBDIR)/checkMetaDb alpha metaDb_${USER} ${DBS}
- create directories under this trackDb directory in the same way as they are under kent/src/hg/makeDb/trackDb, e.g. human/hg19
- don't forget to create empty trackDb.ra files in all these dirs and subdirs
- Now you can cd to usr/trackDb and do a
make update DBS=hg19
- the tracks should now live completely separate in your sandbox and won't be part of the git source tree