Featured Research

New tools for exploring immunogenic neoepitopes in cancer.

Identifying neoepitopes that elicit an adaptive immune response is a major bottleneck to developing personalized cancer immunotherapy therapies. Experimental validation of candidate neoepitopes is extremely resource intensive, and the vast majority of candidates are non-immunogenic, making their identification a needle-in-a-haystack problem. To address this challenge, we developed BigMHC, a deep learning method that predicts MHC-I epitopes and identifies immunogenic neoepitopes with high precision.

View Current Research »


Dr. Rachel Karchin

Professor, Johns Hopkins University. Institute for Computational Medicine, Department of Biomedical Engineering, Department of Oncology, Department of Computer Science.

About the Karchin Lab

We develop computational models to interpret and predict the impact of individual variation in the genome, transcriptome, and proteome. The models are being applied to cancer genomics, unclassified variants in Mendelian disease genes, and complex disease genetics. In collaboration with clinicians, pathologists, and experimental biologists, we aim to make significant improvements in individualized medicine within the next five years.

Lab Members and More Lab Info »


Featured Software Tool Tutorial

Selected Publications

Shao, XM et al. (2022) Ann Oncol. Jul;33(7):728-738 Article

Zheng, L et al. (2022) Bioinformatics Jun 1:btac367 Article

Pagel K et al. (2020) Journal of Clinical Oncology CCI. Mar;4:310-317 Article

View All Publications »

Lab News Archive »

Software and Tools

BigMHC  [github]  [pub]
Predict immunogenic neoepitopes with deep learning 

openCRAVAT  [url]  [github]  [pub]
A modular annotation tool for genomic variants 

PICTograph  [github]  [pub]
Bayesian method to model tumor evolution 

CHASMplus  [github]  [pub]
Cancer-specific driver mutation prediction 

MHCnuggets  [github]  [pub]
Neoantigen prediction with deep learning 

HotMAPS  [github]  [pub]
3D hotspot mutation identification 

20/20+  [github]  [pub]
Cancer driver gene prediction 

SCHISM  [github]  [pub]
Infer clonal evolutionary history of tumors 

VEST  [url]  [pub]
Predict pathogenic variants 

MOCA  [github]  [pub]
Genotype-phenotype correlations 

CLUMP  [github]  [pub]
Cluster Mendelian variants in 1D 

NAPA  [github]  [pub]
Network analysis of multiple mutations 

MuPIT  [url]  [pub]
Visualize mutations on 3D protein structures