Opened 7 years ago

Closed 7 years ago

Last modified 7 years ago

#983 closed task (fixed)

Merge the Tophat alignment and Cufflinks job to a single job

Reported by: Nicklas Nordborg Owned by: Nicklas Nordborg
Priority: critical Milestone: Reggie v4.11
Component: net.sf.basedb.reggie Keywords:
Cc:

Description

This is part of #982.

The Tophat and Cufflinks steps in the secondary analysis should be merged to a single step.

The same item structure should still be created, and protocol/software items should be selected as before.

The sub-folder structure on the file server and in BASE should be the same as before. We do not keep the output from Tophat except log/metadata files that are needed to import the same annotations as before. From the Cufflinks step we need to keep the isoforms.fpkm_tracking file and maybe some more metadata files.

Data from Cufflinks should NOT be imported to the database.

Change History (9)

comment:1 by Nicklas Nordborg, 7 years ago

Status: newassigned

comment:2 by Nicklas Nordborg, 7 years ago

(In [4542]) References #983: Merge the Tophat alignment and Cufflinks job to a single job

Re-ordered things on the Reggie front page and prepared the "Start alignment" wizard by adding settings for the Cufflinks step and the "Confirm cufflinks" wizard by adding information from the alignment step. No backend changes yet.

comment:3 by Nicklas Nordborg, 7 years ago

(In [4543]) References #983: Merge the Tophat alignment and Cufflinks job to a single job

Moved the alignment script generation code to the cufflinks step. The cufflinks step now need a list with MergedSequences items instead of AlignedSequences items.

The scripts have been changed as little as possible and still generate and store the same files as before.

Auto-confirmation has not been changed and may not behave as expected.

comment:4 by Nicklas Nordborg, 7 years ago

(In [4545]) References #983: Merge the Tophat alignment and Cufflinks job to a single job

Changed the script to only keep the isoforms.fpkm_tracking file.

Removed parts of the script that are not needed due to this:

  • picard AddOrReplaceReadGroups
  • fix_cufflinks_tracking_id.sh on the genes.fpkm_tracking and transcripts.gtf files.

No DataFilesFolder annotation on the AlignedSequences item. A side-effect is that no subfolder is created for the alignment step. We get xxx/k.a.t instead of xxx/k.a/t.

Metadata import from the alignment step has been moved to the cufflinks code.

comment:5 by Nicklas Nordborg, 7 years ago

(In [4547]) References #983: Merge the Tophat alignment and Cufflinks job to a single job

Auto-confirm after demuxing is now starting the merged Tophat/Cufflinks analysis.

Auto-confirm rules for alignment has been moved to the cufflinks step.

Removed classes that are not used any longer and started to move away things from the MaskAlignServlet.

comment:6 by Nicklas Nordborg, 7 years ago

(In [4548]) References #983: Merge the Tophat alignment and Cufflinks job to a single job

Registration of failed Tophat/Cufflinks now delete all items up to (including) AlignedSequences.

Removed unused confirmation/registration wizards and counters.

comment:7 by Nicklas Nordborg, 7 years ago

(In [4549]) References #983: Merge the Tophat alignment and Cufflinks job to a single job

More cleanup of unused code.

comment:8 by Nicklas Nordborg, 7 years ago

Resolution: fixed
Status: assignedclosed

Seems to work now.

comment:9 by Nicklas Nordborg, 7 years ago

(In [4566]) References #983: Merge the Tophat alignment and Cufflinks job to a single job

Removing script that is no longer used.

Note: See TracTickets for help on using tickets.