search

Evolinc / Evolinc-I

2 stars 3 forks source link

Questions about Evolinc modifications since publication #17

Open luciaq21 opened 4 years ago

luciaq21 commented 4 years ago

Hi, I have read the paper of the most recent version of Evolinc and I have the following questions:

  1. It says in the paper to run the output FASTA against Rfam, but I see here in the resolved issues github that it says this feature has been added. Do you still suggest that I run my output against Rfam?
  2. Does Evolinc detect only long intergenic non coding RNAs or does it detect other types too?
  3. Since the paper came out, the developers of cuffcompare also made the program gffcompare, which is analogous to cuffcompare and I believe produces the same output. I would like to use gffcompare because its usage is simpler, can I use the output gtf from gffcompare in Evolinc, or do you suggest I stick to cuffcompare? Thank you
andrew-d-l-nelson commented 4 years ago

  1. You are correct - you no longer need to compare your samples against Rfam. We considered that step to be a bit to onerous to have users perform after the fact. Evolinc-I now scans against common contaminants such as ribosomal, the U1-X spliceosomal, and a suite of other housekeeping ncRNAs. It does not remove miRNA precursors and snoRNAs as they have been demonstrated to play roles as lncRNAs.

  2. Evolinc detects intergenic as well as sense and antisense overlapping transcripts. These will be subdivided in the output into separate folders (SOTs and AOTs, respectively). Note that the accuracy of these assignments (sense/antisense) depends heavily on the strandedness of your RNAseq library. As with any interesting discovery, I would suggest RT-PCR, etc., to verify.

  3. GFFcompare is fine - just make sure and use the -r (reference annotation) flag.

Extra: If you run Evolinc on multiple datasets, consider using the "merge_multiple_gtfs.sh" script found in the Accessory Scripts repo. If you do not have cuffmerge in your path, simply swap out instances of "cuffmerge" for "gffcompare" (lines 62 and 72).

Thanks for using Evolinc!