Plasticity can be quantified at single-cell level to gain insight into underlying machinery. This strategy identified FGFR and JAK/STAT inflammation as a driver of plasticity in prostate cancer with combined FGFR and JAK inhibition reversing early stages of plasticity and restoring androgen sensitivity, but only prior to full neuroendocrine transformation.

Our single-cell atlas of SCLC patient tumors identifies plasticity changes in the form of a mixed SCLC-A/N phenotype as well as discrete intratumoral switching between canonical subtypes. We found a recurrent, pro-metastatic PLCG2-high SCLC phenotype that predicts worse overall survival and is associated with a pro-fibrotic myeloid population that is enriched in SCLC vs lung adenocarcinoma.

Our single-cell exploration of castrate-resistant prostate cancer (CRPC) in transgenic mice following RB1, TP53, and PTEN deletion identified multi-lineage infidelity through the switch from a stem-like luminal population to amphicrine, mesenchymal, and multiple neuroendocrine subtypes that show correspondence to neuroendocrine subtypes in small cell lung cancer.

We developed a method based on algebraic topology to model horizontal evolution in viruses and any form of evolution not captured by standard phylogenetics. This method is one of the earliest applications of topological data analysis in the biological sciences.

By developing a method to detect gene fusions in both next-generation whole exome and RNA sequencing, we identified FGFR-TACC fusions in glioblastoma, the first instance of this oncogenic fusion in any cancer type.