search

ambj / MuPeXI

MuPeXI: the mutant peptide extractor and informer, a tool for predicting neo-epitopes from tumor sequencing data.
Other
46 stars 28 forks source link

Will MuPeXI become compatible with NetMHCpan 4.1? #37

Open scyrusm opened 4 years ago

scyrusm commented 4 years ago

I understand that MuPeXI only works with NetMHCpan 4.0. Will it be upgraded in the future, or left as it is?

Thanks!

Pchaskar commented 4 years ago

I have the same question as above mentioned

I managed to run Mupexi with the NetMHCpan 4.1. but I am not very sure about the binding affinity scores

below is an output for the same HLA-peptide using NetMHCpan 4.0 and NetMHCpan 4.1

1) NetMHCpan 4.0 HLA_allele | Chr | Genomic_Position | Mut_MHCrank_BA HLA-A02:01 | 10 | 68742565 | 0.3433

1) NetMHCpan 4.1 HLA_allele | Chr | Genomic_Position | Mut_MHCrank_BA HLA-A02:01 | 10 | 68742565 | 0.5767

I appreciate any help in this regard.

Thanks

worker000000 commented 4 years ago

it is easy to solve this, just revise the origin code in Mupexi.py if not '4.0' in netMHC_path to if (not '4.0' in netMHC_path) and (not '4.1' in netMHC_path)

and the same setting of others

RexTremendae1 commented 4 years ago

You should also parse the output columns since they are different

netMHCpan 4.1 Pos HLA Peptide Core Of Gp Gl Ip Il Icore Identity Score_EL %Rank_EL BindLevel

netMHCpan 4.0 Pos MHC Peptide Of Core Core_Rel Identity Score_EL %Rank_EL Exp_Bind BindLevel

And regarding to Pchaskar' question I would make the neopeptide predictions based on EL rank scores, since netMHCpan was trained with much more EL data than BA, and then calculate binding affinities using NetMHCpan on the reported peptides.

Best

worker000000 commented 4 years ago

so will the output order affect the accuracy of mupexi @RexTremendae1

have you met the following error image

jasonthomson7 commented 2 years ago

@RexTremendae1 thanks for your advice on settiing up NetMHCpan 4.1 with MuPeXI. I'm able to run this now but not sure what you meant when you said we should parse the output columns. Can you please give more detail here? I can't see the columns you references in the Mupexi.py code.

Thanks!

"You should also parse the output columns since they are different

netMHCpan 4.1 Pos HLA Peptide Core Of Gp Gl Ip Il Icore Identity Score_EL %Rank_EL BindLevel

netMHCpan 4.0 Pos MHC Peptide Of Core Core_Rel Identity Score_EL %Rank_EL Exp_Bind BindLevel"