Cancer Drug Target Discovery

Algorithmic Deconstruction of Somatic Copy Number Alterations
for Therapeutic Target Discovery in Cancer

Poster

This project is being conducted in the Beroukhim lab at Harvard Medical School & Broad Institute of MIT and Harvard.

Abstract

Somatic copy number alterations (SCNAs) are among the most prevalent mutational events in the cancer genome, driving oncogenesis by modulating the dosage of key genes. However, the wide-ranging effect of SCNAs often obscures the identification of significant cancer genes, presenting a major challenge in therapeutic target discovery.

Our project aims to address this challenge by developing advanced computer algorithms that leverage complex biological mechanisms underlying SCNA formation. We hypothesize that these efforts will pinpoint genes whose lack or abundance is pivotal to cancer development. With a dataset of over 30,000 tumor samples, our methodology infers haplotype-specific SCNA patterns and reconstructs their duplication and deletion histories. Building upon the biological knowledge of SCNA generation, we then search for genomic regions that exhibit disproportionate alterations, beyond what would be expected if they held neutral effects. This approach allows us to detect genes whose dosage benefits cancer cell proliferation and others that hinder it.

Our result includes four novel algorithms, each addressing a central aspect of this methodology. Preliminary analysis demonstrates their success in identifying candidate regions that appear to be under selection. We anticipate this work to yield critical insights into previously undiscovered targets of SCNAs, and serve as a vital tool for researchers seeking to uncover the underlying cause of their tumor samples. Furthermore, treatment strategies targeting the identified genomic regions will advance personalized therapy options tailored to the unique genetic composition of each patient’s tumor.