Circlehunter: a tool to identify extrachromosomal circular DNA from ATAC-Seq data

Abstract

In cancer, extrachromosomal circular DNA (ecDNA), or megabase-pair amplified circular DNA, plays an essential role in intercellular heterogeneity and tumor cell revolution because of its non-Mendelian inheritance. We developed circlehunter (https://github.com/suda-huanglab/circlehunter), a tool for identifying ecDNA from ATAC-Seq data using the enhanced chromatin accessibility of ecDNA. Using simulated data, we showed that circlehunter has an F1 score of 0.93 at 30× local depth and read lengths as short as 35 bp. Based on 1312 ecDNAs predicted from 94 publicly available datasets of ATAC-Seq assays, we found 37 oncogenes contained in these ecDNAs with amplification characteristics. In small cell lung cancer cell lines, ecDNA containing MYC leads to amplification of MYC and cis-regulates the expression of NEUROD1, resulting in an expression pattern consistent with the NEUROD1 high expression subtype and sensitive to Aurora kinase inhibitors. This showcases that circlehunter could serve as a valuable pipeline for the investigation of tumorigenesis.