266 (PB046) - Evaluating the performance of two NGS-based copy number aberration detection tools in patient-derived organoids

Abstract

Introduction: NGS sequencing technology, including whole genome sequencing (WGS) and whole exome sequencing (WES), has become a primary tool to study genetic variants of tumors. Many tools have been developed for NGS-based copy number aberration (CNA) detection including software GATK (Broad Institute) and CopyWriteR, which both demonstrate reliable performance. GATK software can be applied to WES and WGS data, using target region depth and germline mutation allele frequency for copy number detection. CopyWriteR software uses only WES data, leveraging on the off-target region depth. In this study the advantages and disadvantages of these two tools are compared. Methods: CNA of 23 pairs of patient-derived organoid (PDO) models were analyzed using WGS data with the GATK software, and 18 pairs of PDO models using WES data with both the GATK and CopyWriteR software. CNA was inferred in the presence and absence of paired WGS and WES-based CNA results were compared to evaluate the influence of NGS data type. The WES-based CNA data by GATK and CopyWriteR were compared to evaluate the two algorithms. Results: The number of amplified/deleted regions and genes are comparable for the three combinations (GATK + WGS, GATK + WES, and CopyWriteR + WES) when matched controls were used. WES-based combination detected significantly more CNA events in the absence of matched controls when compared to WGS-based combination. GATK software was prone to unexpected results when the WES capture efficiency was not equal in library preparation. CopyWriteR software missed the presence of some highly amplified genes. Conclusions: WGS data was found to be a superior source for CNA detection when compared to WES data. CopyWriteR software has shortage on detection of high gene amplification events and for WES data, with uneven exon capture rate, it demonstrated superior performance over GATK. These results suggest it is integrating the WES-based results from both software. No conflict of interest.