Publications

This article introduces a multimodal single-nucleus method to profile gene expression and chromatin marks from frozen tumor biopsies, enabling clinical monitoring of non-genetic tumor heterogeneity. In ovarian cancer patients, it reveals consistent evolution under chemotherapy toward resistant EMT/TNFα programs that are epigenetically primed before treatment, and shows baseline tumor states can predict survival.

This AACR 2025 poster  combines single-nucleus RNA and epigenomic profiling from frozen ovarian cancer biopsies to map tumor cell states and track how they evolve under chemotherapy. It identifies 13 recurring tumor phenotypes, shows treatment reduces cycling cells but enriches TNFα/partial mesenchymal states, and links these patterns to survival—supporting single-nucleus profiling as a scalable way to monitor functional tumor heterogeneity in the clinic.

This AACR 2025 poster uses longitudinal single-nucleus RNA sequencing of metastatic urothelial cancer biopsies to understand why many patients stop responding to immune checkpoint inhibitors. It identifies baseline predictors of response and shows that resistance is linked to reduced antigen presentation and interferon signaling, increased T-cell exhaustion, and a shift toward pro-tumoral (M2-like) macrophages.

MAYA is a computational method for single-cell data that detects distinct pathway activation patterns across cell types, rather than a single average score. It is robust to noise and batch effects, can predict cell types from marker genes, and helps identify shared tumor vulnerabilities across patients