William C. Hahn

The paper shows that HSF1 activation supports the emergence of cancer-associated p53 mutations by enhancing their fitness under nutlin-3 pressure.

The paper identifies O-mannosylation and protein maturation as therapeutic targets in BRAF fusion oncogenesis by showing their necessity for fusion-induced transformation.

The paper identifies CREB5 as a driver of immunotherapy resistance in melanoma by promoting collagen matrix deposition, which activates LAIR1 signaling on immune cells.

This study uses deep mutational scanning to analyze KRAS variants, identifying gain- and loss-of-function mechanisms and their impact on oncogenic activity.

The paper introduces a method to compress phenotypic screens using computational deconvolution, demonstrating increased efficiency and applicability to complex multicellular models.

This study identifies PRMT1 as a critical regulator of AR signaling in prostate cancer through genome-scale CRISPR screening and shows its potential as a therapeutic target.

This paper introduces a biologically informed deep learning model for prostate cancer classification and identifies molecular drivers of treatment resistance.

The paper introduces a method to screen large variant libraries with mutations across wide regions using improved mutant detection software, enabling accurate variant calling and counting from sequencing data.

Chronos infers gene knockout fitness effects by explicitly modelling cell proliferation dynamics after knockout, using longitudinal readouts and correcting several known screen biases.

The paper develops a method to predict cell health phenotypes from Cell Painting images using machine learning, validated across multiple perturbations and assays.

The paper uses Perturb-seq to assess the phenotypic impact of somatic variants in cancer, revealing a continuum of gain-of-function KRAS variants.

The paper improves gene dependency estimation from RNAi screens by using model-based normalization and data integration, enhancing accuracy and enabling cross-dataset analysis.

The paper develops CERES, a computational method to correct copy-number effects in CRISPR-Cas9 essentiality screens, improving specificity in cancer cell line studies.

The paper quantifies network mutation burdens across 21 tumor types to identify putative cancer genes, finding 62 new candidates with greater sensitivity for low-frequency mutations.