A complete listing of our manuscripts can be found on Jeff Hammerbacher’s Google Scholar profile. For published papers only, search “J Hammerbacher” on PubMed.


At MUSC, our lab is focused on quantifying T cell biology with a particular interest in understanding how T cells behave in the tumor microenvironment. We use non-animal model systems (Human Primary T Cells: A Practical Guide), genome perturbation (Viable and efficient electroporation-based genetic manipulation of unstimulated human T cells), single-cell measurement (Cytokit: A single-cell analysis toolkit for high dimensional fluorescent microscopy imaging), and probabilistic modeling (Computational and experimental optimization of T cell activation) to interrogate T cell development, diversity, and dysfunction.

We closely collaborate with Paulos Lab, Rubinstein Lab, and Bartee Lab at MUSC to understand what makes T cells more effective for adoptive cell transfer and to explore alternative immunotherapy options.


From 2013 to 2017 our lab at Mount Sinai focused primarily on computational challenges in cancer immunotherapy. An overview of this field can be found in our 2017 Annals of Oncology review paper.


We worked closely with the Bhardwaj lab on the software and methods used in the personalized genome vaccine clinical trial NCT02721043. Together we published a 2015 review paper in Oncology motivating the trial and a 2017 preprint describing the peptide selection pipeline in detail.

We’ve also published preprints describing several components of the PGV pipeline including MHCflurry, our allele-specific Class I peptide/MHC binding predictor, and Vaxrank, our ranking algorithm for candidate vaccine peptides.

Finally, we collaborated with the Bowtell lab on a preprint examining the neoantigen burden of ovarian tumors before and after chemotherapy.


Together with Alex Snyder and Matt Hellmann of Memorial Sloan-Kettering we analyzed data from several clinical trials to discover predictive biomakers for response to checkpoint blockade. In 2017 we published a paper analyzing a cohort of melanoma patients in Cancer Immunology Research and another paper analyzing a cohort of bladder cancer patients in PLoS Medicine.

An important source of biomarkers for checkpoint blockade is the immune contexture of the patient’s tumor. We published a preprint about Infino, our open source software for estimating the immune contexture of a tumor from bulk RNA-seq data. We also blogged about some of the open source software we wrote to facilitate biomarker discovery, including Cohorts and SurvivalStan.


To support our cancer immunotherapy work we wrote a lot of open source software for genomic data processing.

In 2015 we collaborated with a team from Berkeley on a position paper declaring our intent to do data parallel sequence analysis with Apache Spark. Pageant, the Parallel Genomic Analysis Toolkit, is the result of our work in this direction.

To specify and execute complex bioinformatics workflows on shared computing infrastructure we wrote a collection of tools in OCaml that we call the Wobidisco Ecosystem. Also in OCaml we wrote Prohlatype, a sequence-based HLA typer.

For variant annotation, PyEnsembl is a Python interface to the Ensembl database and Varcode computes the impact of genomic variants on protein sequences.

Finally, we were early adopters of typed, packaged, and tested JavaScript for in-browser pileup visualization with pileup.js and we used PEG.js to build an in-browser query language for filtering variant calls in Cycledash.