jlsteenwyk / orthofisher Goto Github PK

Hi Jacob,

Thanks for the nice tool. I was testing orthofisher using a sample dataset with a gene family that we are working with, Auxin Response Factors (ARF), in plants. There are 23 genes in this gene family in Arabidopsis thaliana. I have picked-up three proteins from this family and searched for it in the proteome using HMM that I have generated. As a result I expect to get (ideally) 23 genes or somewhere around or a bit less if the algorithm avoids/filters most remote homologs. But, I get only the three sequences that I gave as input to HMM in the results of orthfisher, indicating that the selection/filtering criteria for orthologs is too stringent. On the other hand, it correctly identifies the three existing orthologs in another proteome, Marchantia polymorpha.

To reproduce the results, I have attached the files that I have used. Could you please check it and resolve this filtering so that multi-copy genes could be estimated properly? Or am I doing something differently?

Looking forward for your response!

Thanks and regards
Sumanth

orthofisher_arf.zip

Hello,

I will be very interested to test your software as an alternative to do reciprocal blast, to recover genes of interest from differents species genomic data to get multifasta of hopefully correct orthologs.

I havent used hmm before. How do we generate the hmm files for the argument -m? If we have a list of genes (gene name or Ensembl ID) can we use this info to fish out in the proteome?

Thanks for the help

regards

Nicolas