no exon lines in the GTF file

[Zeinab5]

Hello,

I got the GTF file from CGD: http://www.candidagenome.org/download/gff/C_albicans_SC5314/Assembly22/ But I got this error message:

Fatal INPUT FILE error, no exon lines in the GTF file: albicans.gtf
Solution: check the formatting of the GTF file, it must contain some lines with exon in the 3rd column.
      Make sure the GTF file is unzipped.
      If exons are marked with a different word, use --sjdbGTFfeatureExon

Is it possible to make STAR accept this GTF file format? or how can I modify this GTF file to be accepted by STAR?

Thanks, Zeinab Hefny

[ewels]

Hi @Zeinab5,

This is a little strange, http://www.candidagenome.org/download/gff/ suggests that the GTF files should use the term exon.

Exactly which GTF file did you use from your link? There are many in that folder.

Phil

[Zeinab5]

This is the one that I used http://www.candidagenome.org/download/gff/C_albicans_SC5314/Assembly22/C_albicans_SC5314_A22_current_features.gtf

[varemo]

The corresponding GFF file (http://www.candidagenome.org/download/gff/C_albicans_SC5314/Assembly22/C_albicans_SC5314_A22_current_features.gff) seems to contain the required fields, as oposed to the GTF which only contains "CDS". Is it possible to supply a GFF file directly to the pipeline?

[ewels]

I think it should be fine to run with a GFF instead of GTF. The format is pretty similar right? Certainly easy enough to try... Just use --gtf [path]

[Zeinab5]

Hi, I solved the issue by converting the GFF file from the CGD to a GTF file by using gffread and it worked. I couldn't use the gff file directly with the pipline because gtfToGenePred can't convert gff to GenePred which I need to get the Bed12 file. it also didn't work when I didn't gave the path of the Bed12 to the pipline.

[ewels]

I solved the issue by converting to the GFF file from the CGD to a GTF file by using gffread and it worked.

Ok great! You mean this tool right? Maybe we can bundle that with the pipeline and add an additional -gff option...

It also didn't work when I didn't gave the path of the Bed12 to the pipeline

You mean something didn't work with --bed12? What was the error?

[Zeinab5]

Yes, this is the tool that I used.

./gffread my.gff3 -T -o my.gtf

I thought that was the reason, attached the error message

[ewels]

Hmm, Non existent bind point usually means that you don't have overlays enabled in singularity. This is to do with your singularity installation - if you have an old version of the linux kernel then it may be impossible. If this is the case, then the singularity container we provide won't work for you.

The rest of the error is not so clear. But you don't get any errors now, running with this GTF file?

[Zeinab5]

Yes, It's working fine withe this GTF file now

[ewels]

Ok great - then maybe the bind point thing was a false alarm. Strange, I'll give it another test here to be sure.

I've created a new issue for GFF support, but over at nf-core/RNAseq (we are porting our pipelines to this new nf-core organisation - see http://nf-co.re/). Issue here: https://github.com/nf-core/RNAseq/issues/15

[Zeinab5]

ok, Nice. thanks for the links