Open nick-youngblut opened 3 years ago
I am running barnap on 100's of genomes, and this error is only occurring for a few of them, which makes it seem to be input-specific, but the genome files appear to be valid.
Is the error caused by a lack of usable output by nhmmer?
It's been more than a year since Nick reported this issue and I recently encountered it too.
After hours of wondering i finally looked into the tool perl code and found why; it's simply because of line 99 in the executable (in bin dir): err("nhmmer failed to run - $_") if m/fail|error|core dump|bus error/i;
In Nick's first caption, the file path contained the fail word - X361_fail_Common_Opossum__maxbin2__High.001.fna
And that perl line looks for that and therefore thinks that it is a fail error message from nhmmer...
So changing just the directories or the files name that barrnap is using to avoid any of those key words used in line 99 regex in the path is going to fix the error. Definitely something that can be improved easily if a Barrnap.10 is coming
@greboul: Thanks, that helped me figure out my problem too! In my case this error occurs whenever the infput fasta files contain the substring "error" somewhere in the filename. Weird bug but luckily fixable if one knows the problem...
I'm getting the following error:
However, when I activate that conda env and run nhmmer myself, the run completes successfully:
The input genome fasta file contains 123 contigs, and it is a valid fasta file.
My conda env: