Computational Spatial Modeling of Extrachromosomal DNA to Decipher Cancer Evolution.

Imaging techniques such as DNA fluorescent in situ hybridization (FISH) and nascent RNAscope enable detection and quantification of extrachromosomal DNA (ecDNA) in tumor cells. Here, we describe a workflow of ecDNA detection and spatial characterization using DNA FISH, and integration of these data with a spatial evolutionary model to estimate past evolution in individual patients' tumors, such as brain cancers. This analysis leverages the spatial patterns of ecDNA measured across multiple spatial regions of a solid tumor to predict the competitive advantage conferred by oncogenic ecDNA to their host tumor cell, and the number of these ecDNAs present around the time of the beginning of clonal expansion of the tumor, informing us on the origins of cancer.