MHCnuggets @ Karchin lab

MHCnuggets

predicts the binding affinity of peptides to class I and class II Major Histocompatibility Complex (MHC) molecules

Binding of peptides to Major Histocompatibility Complex (MHC) proteins is a critical step in immune response. Peptides bound to MHCs are recognized by CD8+ (MHC Class I) and CD4+ (MHC Class II) T-cells. Successful prediction of which peptides will bind to specific MHC alleles would benefit many cancer immunotherapy appications. MHCnuggets provides a fast, accurate, and unified framework for the prediction of peptide-MHC binding for both class I and class II alleles. This includes rare alleles i.e. ones not well represented in its training data. It does so by taking advantage of recent advances in deep learning. Concretely, MHCnuggets is a collection of long short-term memory networks trained using a transfer learning protocol. This helps MHCnuggets deal with two of the primary challenges faced in neoantigen prediction today -- variability of peptide length, and paucity of data in rare alleles.

MHCnuggets is free for non-commercial use. For more details please refer to our Software License. Commercial users should contact the Johns Hopkins Technology Transfer office.

Current stable release is mhcnuggets-2.1 last updated on 06/29/2018.

Source Code Releases

You can view the current source code on github.

mhcnuggets-2.1.tar.gz    06/29/2018    Latest stable release

Workflow

Please refer to the user guide for a quick start to installing and running MHCnuggets!

Platform

MHCnuggets is pip installable on any platform running python 2.7 or 3.6. Find more details on installation here.

Primary citation

If you use our software for a publication, please cite the following:

Bhattacharya R, Sivakumar A, Tokheim C, Beleva Guthrie V, Anagnostou V, Velculescu VE, Karchin R (2017) Evaluation of machine learning methods to predict peptide binding to MHC Class I proteins. Submitted [bioRxiv preprint]

Software primary contact/developer

Rohit Bhattacharya:  rbhatta8 at jhu dot edu