Creating a pushQ Entry
After you have finished creating a track and it is ready for QA, you should add an entry to the qapushq table on hgwbeta. There is a CGI for this.
Add a track to the pushQ only if everything is ready. This includes:
- All tables have been built.
- TrackDb has been edited (including search stanzas). If this is an update to an existing track, you may need to use release tags.
- Make alpha on hgwdev in src/hg/makeDb/trackDb directory.
- All underlying files are complete and in the correct location (usually /gbdb/).
- All associated downloads files have been created and are in the correct location (usually /data/apache/htdocs-hgdownload/goldenPath/), and accompanying README.txt and md5sum.txt files are present.
- html text for details page is complete (see also description pages and citation formats) and is in highest logical level for the assembly.
- All tables have been added to the all.joiner file.
- Data contributor (if any) has approved of the track display.
Editing the pushQ entry:
PRIORITY: Leave set to A.
IMPORTANCE: Leave set to Unprioritized.
NEW TRACK?: If this same track exists on this or another assembly, or this is an addition to an existing track, leave this set to N. If this is a brand new track set this to Y. If this isn't a track at all, set this to X.
TRACK: Put an identifier for the track here. This is often the shortLabel.
RELEASE LOG: Generally you should put the track shortLabel followed by comments in parentheses, if any (e.g. Known Gene (data update)). No need to enter the database name. This text will appear in the Release Log.
RELEASE LOG URL If you would like the Release Log entry for this track to link to the track description page, then enter a relative link to it here, e.g.:
../../cgi-bin/hgTrackUi?db=hg18&g=wgEncodeBuOrchid
DATABASES: List the database(s) where your table(s) are located (e.g. hg19, hgFixed).
TABLES: List all tables that should be pushed. Don't forget to include:
- extFile
- seq
CGIs: If this track depends on a particular upcoming release of the CGIs, enter the CGI name and the version number here (e.g. hgTracks (v250)).
FILES: List all files that should be pushed. Don't forget to include:
- gbdb files
- download files
CURRENT LOCATION: Set to hgwdev.
SIZE: Press the Show Sizes button. All tables and files should be found without errors (if they are not, you probably have entered a path incorrectly or have a typo in the lists). Scroll to the bottom of the page and click the "Set Size as ..." link.
MAKEDOC VERIFIED: Leave this set to N.
INDEX VERIFIED: Leave this set to N.
ALL.JOINER VERIFIED: Leave this set to N.
EXTERNAL SOURCE: If applicable, enter the name and email address of the collaborator, or enter a website or lab where the data came from.
NOTES: Put the redmine issue number here. In the redmine ticket, add:
- Any notes you think might be helpful to the QA team. Be as specific as possible -- the more you say, the less QA will have to ask you.
- If this track has been requested by a user, add a note with the user's email address and a link to the redmine MLQ ticket, if there is one. QA will send an email to that user after the track has been pushed to the RR.
- If this track is replacing an existing track, list the tables/files to drop.
After the track has been QAed and pushed to the public website, you can edit the original tables and files on genome-test as you like. This new data will not be QAed and pushed until you add another pushQ entry. If you edite the trackDb entry, though, be sure to use release tags!!!
If this is the release of an entire assembly, use Angie's makePushQSql.pl script.
