NBISweden / AGAT

Another Gtf/Gff Analysis Toolkit
GNU General Public License v3.0
467 stars 56 forks source link

agat_sp_fix_overlaping_genes.pl does not seem to complete #359

Closed sanyalab closed 1 year ago

sanyalab commented 1 year ago

Describe the bug agat_sp_fix_overlaping_genes.pl starts running fine, and parses the gff3 input file.

=> OmniscientI total time: 125 seconds isoseq_gmap_clean_1.gff file parsed

After this there is no output on screen. Blank lines or carriage returns are printed on screen. The tool never finishes. I have to use Cntrl-C to kill. No output file is produced.

General (please complete the following information):

To Reproduce agat_sp_fix_overlaping_genes.pl -f isoseq_gmap_clean_1.gff -o isoseq_gmap_clean_1_out.gff

The input file size is more than 25 MB therefore cannot upload.

Expected behavior A clear and concise description of what you expected to happen.

Screenshots isoseq_gmap_clean_1_RunOutput

sanyalab commented 1 year ago

Hi Jacques, I used verbose. Seems to be working fine. Will restart this thread if I notice errors. Thanks Abhijit

Juke34 commented 1 year ago

You may also try a more recent AGAT version. I think this script has been updated.

sanyalab commented 1 year ago

Hi Jacques,

The new version is definitely faster than the previous. This script worked fine.

I have a query. I have several genome annotations for the same line achieved using a Liftover strategy. I want to consolidate these annotations into a single one by getting rid of identical isoforms, overlaps, merging between isoforms. Is there any suitable tool to do this in the AGAT suite. I used fix_overlap, but it still gives me a lot more genes. Any advice?

Thanks Abhijit

Juke34 commented 1 year ago

Hi, I guess the _merge_annotation is the one you need to achieve that.

sanyalab commented 1 year ago

Hi Jacques,

Whats the difference between merge_annotations, and complement_annotations? In both cases do identical isoforms get removed? I want to the final annotation to retain non-identical isoforms per loci.

Thanks Abhijit