The MSK Genomic Data Analysis Center for Tumor Evolution

Abstract The MSK Genomic Data Analysis Center for Tumor Evolution seeks to implement tools, best practices and analytical workflows for studying cancer evolution from cancer genome and transcriptome sequencing data. Over the last 15-years, survey sequencing of patient populations of many cancer types has elucidated novel

driver mutations which are mechanistically responsible for disease pathogenesis. The Cancer Genome Atlas (TCGA) and individual laboratory efforts have broadened the understanding of biological processes impacted by somatic mutation and revealed new therapeutic targets that have achieved clinical impact. However, most of this

work has been based on bulk DNA sequencing from primary tumors and single biopsies from patients. It is well understood that cancer is an evolutionary process during which clonal expansions within patients generates heterogeneity and phenotypic diversity of cell populations across metastatic sites over time (with or without

therapeutic intervention). Indeed, the same targeted therapies developed based on mutation discoveries often select for resistant clones, keeping durable cures out of reach. We will develop analytical methods, tools and software infrastructure to study cancer progression through the lens of evolution, shifting emphasis from analysis

of primary tumors to dynamic analyses over clinical trajectories. We expect our program will advance the ability to study clinical trajectories of patients in a more comprehensive approach, including temporal, spatial and single cell analysis to better represent the full clonal repertoires of tumors and to study the determinants of how and

why tumors evolve. We use tools, well established in our laboratories, in three key areas: i) variant interpretation from metastatic and post-treatment samples for discovery of therapeutic resistance mutations (Aim 1); ii) multi- sample analysis across anatomic space, and/or time series data from serial biopsy or cell free DNA to track and

model clonal dynamics (Aim 2); iii) single cell approaches for clonal decomposition and clone-specific phenotyping within patients (Aim 3). Our team is well positioned to carry out our objectives having developed leading software infrastructures supporting TCGA and clinical sequencing through MSK-IMPACT, development

of clinically approved assays for longitudinal monitoring of patients through cell free DNA sequencing (MSK- ACCESS) and through study of clonal evolution at bulk and single cell resolution. We will implement and improve tools to support each of these aims, including Cancer Hotspots, OncoKB, and cBioPortal for Aim 1, PyClone

and fitClone for Aim 2 and CloneAlign and CellAssign for Aim 3, tailoring and customizing software to support investigations into the dynamic and evolutionary nature of human cancers. These tools comprise a software infrastructure focused on cancer evolution through variant allele interpretation, multi-sample analysis and single

cell investigation. Our infrastructure will enable researchers to automate evolutionary interpretation of disease dynamics to better understand the clinical end points of metastatic progression and therapeutic resistance.