Peptide-MHC I binding affinity prediction
Python Shell
Latest commit 44bd56e Jun 28, 2018

Build Status


MHC I ligand prediction package with competitive accuracy and a fast and documented implementation.

MHCflurry supports Class I peptide/MHC binding affinity prediction using ensembles of allele-specific models. It runs on Python 2.7 and 3.4+ using the keras neural network library. It exposes command-line and Python library interfaces.

If you find MHCflurry useful in your research please cite:

T. J. O’Donnell, A. Rubinsteyn, M. Bonsack, A. B. Riemer, U. Laserson, and J. Hammerbacher, "MHCflurry: Open-Source Class I MHC Binding Affinity Prediction," Cell Systems, 2018. Available at:

Installation (pip)

Install the package:

$ pip install mhcflurry

Then download our datasets and trained models:

$ mhcflurry-downloads fetch

You can now generate predictions:

$ mhcflurry-predict \
       --alleles HLA-A0201 HLA-A0301 \
       --out /tmp/predictions.csv
Wrote: /tmp/predictions.csv

See the documentation for more details.

MHCflurry model variants and mass spec

The default MHCflurry models are trained on affinity measurements. Mass spec datasets are incorporated only in the model selection step. We also release experimental predictors whose training data directly includes mass spec. To download these predictors, run:

$ mhcflurry-downloads fetch models_class1_trained_with_mass_spec

and then to make them used by default:

$ export MHCFLURRY_DEFAULT_CLASS1_MODELS="$(mhcflurry-downloads path models_class1_trained_with_mass_spec)/models"

We also release predictors that do not use mass spec datasets at all. To use these predictors, run:

$ mhcflurry-downloads fetch models_class1_selected_no_mass_spec
export MHCFLURRY_DEFAULT_CLASS1_MODELS="$(mhcflurry-downloads path models_class1_selected_no_mass_spec)/models"