Research

The Beroukhim Lab’s work focuses on understanding the genetic changes that occur through cancer evolution and how these changes affect cancer behavior. We have particular interests in brain cancers and in alterations of chromosome structure (rearrangements and copy number changes) across many cancers. We have also had longstanding efforts in endometrial cancers.

We have research activities in the following areas:

Developing and applying computational methods to analyze large genomic datasets. We have developed computational approaches to detect copy-number alterations and rearrangements from high-throughput sequencing data, to distinguish positively selected from passenger events, and to determine functionally relevant associations between genetic events. We have applied these approaches to genomic data representing tens of thousands of cancers, determining mechanistic and selective forces shaping the cancer genome.

Identifying novel cancer vulnerabilities associated with copy-number change. Through integrated analyses of genetic and functional genomic data, we were the first to detect cancer vulnerabilities associated with copy-number losses (and invented an extremely convoluted acronym for them: CYCLOPS, standing for “Copy-number alterations Yielding Cancer Liabilities Owing to Partial losS”). We have shown that this is the most enriched class of dependencies associated with copy-number changes, and in one case, have identified a small molecule that exploits a CYCLOPS dependency, indicating a potentially generalizable approach to translating these vulnerabilities into treatments.

Performing comprehensive genomic analyses of brain cancers in both adults and children. We identified SMO and AKT1 mutations in meningiomas, leading to international clinical trials of SMO and AKT inhibitors in patients with these tumors. We also identified MYB-QKI rearrangements as pathognomonic of angiocentric gliomas in children, and showed that these rearrangements simultaneously act through at least three direct mechanisms to generate tumors—the most complex effects yet shown for an oncogenic rearrangement. We are also performing comprehensive small molecule, functional genomic, and epigenomic profiling to determine the effects of and vulnerabilities associated with genomic alterations across a range of brain tumors.

Studying tumor evolution. We have characterized genomic differences between primary and metastatic cancers across a range of cancer types, evaluating tumor heterogeneity through space and time. We developed new analytic approaches to show that, for endometrial cancers, metastases tend to arise from a single clone within the primary tumor. We also have made extensive efforts to characterize the changes that brain tumors undergo through conventional and novel treatments to probe mechanisms of treatment resistance and how to overcome them. These efforts include single-cell DNA and RNA sequencing and development of new tools, such as novel cell barcoding approaches to dissect the evolution of cell populations through time.