1. Casey Dunn
  2. agalma
  3. Issues
Issue #42 resolved

Add stats on chimeric transcripts

Casey Dunn
repo owner created an issue

Ya and Smith did a detailed analysis of chimeras generated during transcriptome assembly:

[http://www.biomedcentral.com/1471-2164/14/328]

We should incorporate their script for identifying chimeras from blastx results into postassemble.py. It would run right after the nr_annotate stage. The fraction of each type of chimera would be recorded and displayed in the transcriptome and tabular reports.

We could optionally exclude transcripts that are found to be chimeric from future analyses.

Comments (3)

  1. Casey Dunn reporter

    Their findings could also be used to improve exemplar selection. In particular, we should use their pickH strategy rather than merge paths as we currently do. This would be done in assemble.py.

    Their blastx approach to chimera detection could still be applied in postassemble.py - these are two independent analyses.

  2. Log in to comment