This is a test version of Biostars. For the public version, visit https://www.biostars.org.
BUSCO IO error on trans mode: '[Errno 2] No such file or directory'

Hi, I am comparing transcriptome assemblies from a few different assemblers and have been using BUSCO (http://busco.ezlab.org/) to characterize the completeness of assembly compared to a set of orthologues.

All seems to work well for some transcriptome assemblies (e.g. Trinity) but not for others (Oases) - I get this error:

*** Extracting candidate transcripts! ***
Traceback (most recent call last):
  File "/root/Downloads/BUSCO_v1.22/BUSCO_v1.22.py", line 565, in <module>
    out = open('%s%s%s_.temp' % (args['tmp'],i,args['abrev']),'w')
IOError: [Errno 2] No such file or directory: './Locus_1_Transcript_5/10_Confidence_0.455_Length_2461new_.temp'

after running this

root@:~  python /root/Downloads/BUSCO_v1.22/BUSCO_v1.22.py -o new -in  
    ./transcripts-l200.fasta -l /root/Downloads/BUSCO_v1.22/metazoa/ -m trans

All dependencies installed, and I as I said, works for some fasta assemblies but not others. I thought it might be issues with the long fasta headers, but that error does not suggest such a problem. Looking at that line in the python code;

          if i in scaff_list:
   565    out = open('%s%s%s_.temp' % (args['tmp'],i,args['abrev']),'w')
            out.write('>%s\n' % (i))
            check = 1

Any ideas?

Thanks for any help.

LP

rna-seq busco linux

1 answer

Actually, it was the headers. There is a forward slash in each of the fasta headers in the Oases assembly output - that was the offender. Replacing "/" with an underscore character in all headers solved the problem.

Hello,

The new version of BUSCO that has just been released (in beta when I write this comment) is better at dealing with that kind of issues. It does not crash and warn you, if there is an issue with your header.

You will find it on http://busco.ezlab.org/v2/ and https://gitlab.com/ezlab/busco

And from now, you can get a quick response on gitlab about BUSCO related issues.

Mathieu

Log in to answer this question.