BioPhi icon indicating copy to clipboard operation
BioPhi copied to clipboard

Published 20 hours ago • verified publisher icon Merck

Problems about human germline sequence

Open LLei2 opened this issue 2 years ago • 4 comments

While using humanization module at https://biophi.dichlab.org/, I found the sequence of IGKV2-30*01 was different from the sequence from http://www.imgt.org/download/GENE-DB/IMGTGENEDB-ReferenceSequences.fasta-AA-WithoutGaps-F+ORF+inframeP .The second sequence was copied from the biophi web page. image image

Maybe there are some errors with your human germline V and J gene sequences.

LLei2 avatar Aug 18 '21 07:08 LLei2

Hi @LLei2, thanks for reporting this. Looks like an issue with ANARCI numbering, our IGKV2-30*01 sequence is identical to GENE-DB but it gets scrambled during grafting. We'll look into this.

prihoda avatar Aug 18 '21 08:08 prihoda

Ok, this seems like an ANARCI issue indeed. I fixed it on the public server by installing an older version of ANARCI: conda install anarci==2020.04.23. Can you try submitting at https://biophi.dichlab.org/ again please?

We'll need to submit an issue to the ANARCI repo, not sure what change introduced this error.

prihoda avatar Aug 18 '21 18:08 prihoda

After resubmitting, the returned results make sense. Thanks for fixing this issue.

LLei2 avatar Aug 19 '21 00:08 LLei2

Great, thanks. Closing this now, since it seems like an ANARCI issue. I submitted an issue report to the ANARCI repo: https://github.com/oxpig/ANARCI/issues/17

BioPhi will now require ANARCI version from 2020 until this is resolved.

prihoda avatar Aug 21 '21 11:08 prihoda