Merck / deepbgc

BGC Detection and Classification Using Deep Learning
https://doi.org/10.1093/nar/gkz654
MIT License
127 stars 27 forks source link

different number of elements in *.bgc.tsv rows #49

Closed artur-matysik closed 3 years ago

artur-matysik commented 3 years ago

Hi,

recently I encounter the problem of different number of elements in the*.bgc.tsv output files. I am not sure if it is input specific or just result of the update to 1.25. For some rows, there seems to be missing '\t's (product_activity variable), so the columns are shifted and can't be correctly imported as TSV. Attached is the output I got and the.gbk that produced the error.

Thanks for having a look!

deepBGC_issue.zip

prihoda commented 3 years ago

Hi @artur-matysik, thanks for reporting again. This was caused by cluster annotations already being present (from antiSMASH), which was handled incorrectly in our TSV export function. You can pip install deepbgc==0.1.26. I also included some dependency changes that I had been wanting to fix for some time, so hopefully it didn't break anything.

artur-matysik commented 3 years ago

Works like a charm! Thanks @prihoda : )