Abstract Cancer is an evolutionary process, in that a tumor results from a series of genomic mutations that are acquired over time. The history of how a tumor evolved is often represented using a clonal tree: a particular type of rooted tree where vertices represent tumor cell clones and edges represent ancestral relationships between them. In recent years, numerous algorithms have been developed that infer clonal trees from sequencing data. Correspondingly, researchers have also developed distance measures to compare clonal trees. While such distance measures have proven valuable, they are limited in a variety of ways. First off, they usually only utilize one or a few types of specific information when being computed - making it challenging to assess how well they capture other common types of differences between clonal trees. Secondly, their final output is just a single numerical value - meaning that information on what parts of the trees contributed to the distance is not being readily conveyed. In this work, we address both of these issues. We define a set of four classes of tree differences that describe common ways that clonal trees might differ from each other and analyze these classes using three different clonal tree distance measures. We then introduce a visualization tool called VECTr ( V isual E ncodings for C lonal Tr ees) that provides an enhanced pairwise comparison of clonal trees using the same three distance measures. In particular, VECTr contains three types of visualizations that allow a user to easily identify what parts of a tree contribute to its distance, including information relevant to when the trees differ in a manner that corresponds to one of our four classes of tree differences. We demonstrate the utility of our visualization tool using a variety of different use cases, including application to both simulated and real clonal trees. Code associated with the project is freely available at: https://github.com/quocodile/visualize-clonal-trees CCS Concepts Applied computing Molecular evolution; Computational genomics; Computational biology .
Traw et al. (Wed,) studied this question.