A simple tool for using NetMHCpan and NetMHCIIpan using multiple CPUs in a Python environment.
Usage:
from PyNetMHCpan import NetMHCpan
h = NetMHCpan.Helper(mhc_class='I',
alleles=['HLA-A03:02', 'HLA-B07:02'],
n_threads=12)
# If the configuration needs to be updates, a prompt should appear (i.e. you need to set the paths
# to NetMHCpan/NetMHCIIpan).
# If it is the first time you have run the program and a prompt did not appear, you should
# run the following line (as it is incredibly unlikely that the default paths to NetMHCpan and
# NetMHCIIpan are correct for your system)
h.update_config()
# A prompt will appear showing your the current configuration and asking if you would like to update.
# Note that if you only use one of the tools (i.e. NetMHCpan or NetMHCIIpan) you can leave the path to
# the other as it is or change it to something nonsensical. The program will only check the path of the tool
# it needs (according to the mhc_class parameter).
h.add_peptides(['N.ENSDESYSEEEEEMPDSD.E',
'F.EC[+57]NNSESGYLYC[+57]C[+57]GTC[+57]Y.Y',
'K.AM[+15]IM[+15]C[+57]QGC[+57]GAFC[+57]HDDC[+57]IG.P'])
# Note that there is no need to remove modifications or previous/next amino acids from your
# search engine output! That is automatically taken care of.
# now run the predictor.
h.make_predictions()
# Once it is done, access them as follows:
h.predictions
# returns a pandas dataframe:
# Peptide Allele Rank Binder
# 0 ECNNSESGYLYCCGTCY HLA-A03:02 85.000 Non-binder
# 1 ENSDESYSEEEEEMPDSD HLA-A03:02 100.000 Non-binder
# 2 AMIMCQGCGAFCHDDCIG HLA-A03:02 100.000 Non-binder
# 3 ECNNSESGYLYCCGTCY HLA-B07:02 92.500 Non-binder
# 4 ENSDESYSEEEEEMPDSD HLA-B07:02 100.000 Non-binder
# 5 AMIMCQGCGAFCHDDCIG HLA-B07:02 100.000 Non-binder
# Note that the dataframe is in long format. If you have multiple alleles, there will be multiple
# obeservations for each peptide (one per allele).The binding cutoffs for the Binder field are as follows:
| MHC class | Strong binders | Weak binders |
|---|---|---|
| I | 0.5 | 2.0 |
| II | 2.0 | 10.0 |
For more information on NetMHCpan and NetMHCIIpan or to download them see the following:
- https://services.healthtech.dtu.dk/service.php?NetMHCpan-4.1
- https://services.healthtech.dtu.dk/service.php?NetMHCIIpan-4.0
Note that these tools only run on Linux and MacOS machines. On Windows you can set up the Windows Subsystem for Linux, which will let you run a Linux terminal! I haven't tried it out with PyNetMHCpan, but I assume so long are you are also running it from the Linux terminal it should work.