MakePushQSql.pl: Difference between revisions

From genomewiki
Jump to navigationJump to search
mNo edit summary
(Replacing page with 'This page is no longer maintained.')
 
(One intermediate revision by one other user not shown)
Line 1: Line 1:
This script was written by [[AngieHinrichs | Angie]]. 
This page is no longer maintained.
 
It is a script which assists developers in creating a [http://hgwbeta.cse.ucsc.edu/cgi-bin/qaPushQ qapushq table] for a
new genome db's push queue. 
 
Suggested usage:  '''makePushQSql.pl $db > ~/qapushq.$db.sql'''
 
It prints SQL commands (that look a lot like hgsqldump output) to
stdout.  It's best if you redirect the output to a file, edit it to
make sure it's all correct, and then hgsql it on hgwbeta.  After that
point, use the web Push Queue tool to make any further changes.
 
When it can't figure out where a table belongs, or can't find an
expected download file, it prints a message to stderr.  When done, it
prints more editing suggestions.
 
It uses trackDb and the joinableFields program to guess which tables
belong to which tracks.  However, sometimes there are a couple levels
of indirection in all.joiner, and sometimes a table might be missing
from all.joiner, but again it will alert you.
 
It does not look very hard for /gbdb/ files that may be associated
with a track, but it provides some SQL commands to run on hgwdev to
list extFile and wiggle table files in $db, and then you can edit
those in.
 
Other known limitations:
* no concept of composite tracks (but I think those are still somewhat rare in initial releases)
* does not look for goldenPath/ stuff aside from bigZips, database and chromosomes -- I might add something for chains and nets.
 
This script also makes a single entry in the main pushQ table, which alerts
QA to the fact that there is a new assembly sub-pushQ.
 
 
[[Category:Browser QA]]

Latest revision as of 19:15, 10 March 2011

This page is no longer maintained.