Augustus output (GFF file) to Anvi`o

Question

[zion22]

Hi, thanks for reading this. I'm very amateur in this area and I have a question, I am using Augustus to predict genes of a eukaryotic organism and I would like to integrate the output of Agustus (gff file) to the contigs file of Anvi´o. I tried to use GFFPraser but couldn't.

So if anyone could advise me what to do I would be very grateful.

[Juke34]

Do you use the option —gff3 in Augustus?

Because is you use the GTF output it is a bit wierd (at least for the versions I have used). It doesn't follow the specifications. First they use features types (3rd column) that are not allowed. This is fine in mos of cases, I’m mean it’s rarely a problem because the tools tend to skip non expected features. The main problem is the top feature I.e gene, that have just a value (name/id) in the 9th column while it must be tag value where the tag must be ID.

Anyway if you use the gff3 option it should be better. The only problem is that they use features types (3rd column) that are not allowed by the specifications. But it shouldn't be a problem anyway.

To make it standard and pinpoint potential problems, you can have a try of the script gxf_to_gff3.pl from the GAAS repository using the verbose option (-v 1).

Answer 1

[Philipp Bayer]

I've made good experiences with this script: https://github.com/jorvis/biocode/blob/master/gff/convert_augustus_to_gff3.py when downstream software wasn't liking the AUGUSTUS gff3 (you'll have to run pip install biocode first)