Extract transcripts from GTF to BED12

Ralf's Perl script is old and not very general - it does not work on old GTFs. Should we substitute it?

added Discuss label

unassigned @kanitz

Yes, lets use some python gtf/gff parser instead (e.g. gff reader from HTSeq). An example of using it we have here.

assigned to @gypas

@bakma Can you please point me to some test files of the perl script?

@gypas notice the call in that rule:

        "(gtf_transcript_type_to_bed12.pl \
        --anno={input.gtf} \
        --type=protein_coding > {output.bed12}); \
        2> {log.stderr}"

You only have to provide a GTF. I have been always running it with the full one, it does not take much time. But I guess you could provide just chr21 as well.

changed milestone to %v0.2.0 release

added To Do label

added Doing label and removed To Do label

I updated the rule for the conversion in the branch gtf2bed_update but it seems that we do not have any tests for the tin score calculation and the corresponding files.

@bakma Is this the case or am I missing something?

We have tests for all three tin-related scripts in the tin score repo:
https://git.scicore.unibas.ch/zavolan_group/tools/tin_score_calculation
under tests directory

Thanks, but these are not exactly tests. These are just triggering the execution of the script. The same happens for https://git.scicore.unibas.ch/zavolan_group/tools/gtf_transcript_type_to_bed12.

In any case, the new script is here: https://github.com/zavolanlab/zgtf. It would be nice if you can test it and report any issues.

You can run it via singularity as well:

singularity pull docker://zavolab/zgtf:0.1
singularity exec zgtf gtf2bed12 --help

mentioned in merge request !65 (merged)

Merged in !65 (merged)

closed

added 1 deleted label and removed Doing label