UCL Computational Cancer Seminar Series: Professor Javier Herrero

Integrating copy number aberrations in tumour phylogenetic analyses

Professor Javier Herrero, Head of the Bill Lyons Informatics Center (BLIC), UCL Cancer Institute.

Abstract:

The phylogenetic analysis of a tumour is possible using multiple related samples to accurately disentangle clonal compositions and infer the most likely evolutionary tree. With a few notable exceptions, most tools for tumour phylogenetic inference have relied on somatic point mutation data. However, copy number aberrations (CNAs) are one of the main drivers of tumours heterogeneity. While whole-genome or whole-exome sequencing approaches are ideal for detecting copy-number events, DNA methylation arrays are routinely used in cancer genomics, for instance for the classification of central nervous system tumours. In addition to studying differences in DNA methylation, these arrays can also be used to estimate CNAs.

We present MultiSEPhyR (https://github.com/chulingding/MultiSEPhyR), a novel computational tool for deconvolving copy number events based on multi-sample methylation array data and its application to the GCGR cohort, where we compare 56 pairs of GSC lines with their primary glioblastoma and with 26 xenograft-rederived cell lines from 9 of these cases.

With the advent of CNA-based phylogenetic inference tools like MultiSEPhyR and HATCHet, it is possible to infer phylogenetic trees based on either somatic mutations or somatic CNAs. However, combining mutational and copy number data to provide an integrated perspective of the evolution of a tumour requires new approaches. We have developed MAPSCE (MAPping Sub-Clonal Events; https://github.com/MarkTranHS/MAPSCE), a tool that allows mapping sub-clonal events in multi-sample multi-omics datasets. We show how MAPSCE can be used to map CNA on phylogetic trees based mutational data, providing a shared framework to study the co-evolution of CNA and point mutations.