Hierarchical discovery of large-scale and focal copy number alterations in low-coverage cancer genomes

Abstract

Background: Detection of DNA copy number alterations (CNAs) is critical to understand genetic diversity, genome evolution and pathological conditions such as cancer. Cancer genomes are plagued with widespread multi-level structural aberrations of chromosomes that pose challenges to discover CNAs of different length scales, and distinct biological origins and functions. Although several computational tools are available to identify CNAs using read depth (RD) signal, they fail to distinguish between large-scale and focal alterations due to inaccurate modeling of the RD signal of cancer genomes. Additionally, RD signal is affected by overdispersion-driven biases at low coverage, which significantly inflate false detection of CNA regions. Results: We have developed CNAtra framework to hierarchically discover and classify 'large-scale' and 'focal' copy number gain/loss from a single whole-genome sequencing (WGS) sample. CNAtra first utilizes a multimodal-based distribution to estimate the copy number (CN) reference from the complex RD profile of the cancer genome. We implemented Savitzky-Golay smoothing filter and Modified Varri segmentation to capture the change points of the RD signal. We then developed a CN state-driven merging algorithm to identify the large segments with distinct copy numbers. Next, we identified focal alterations in each large segment using coveragebased thresholding to mitigate the adverse effects of signal variations. Using cancer cell lines and patient datasets, we confirmed CNAtra's ability to detect and distinguish the segmental aneuploidies and focal alterations. We used realistic simulated data for benchmarking the performance of CNAtra against other singlesample detection tools, where we artificially introduced CNAs in the original cancer profiles. We found that CNAtra is superior in terms of precision, recall and f-measure. CNAtra shows the highest sensitivity of 93 and 97% for detecting large-scale and focal alterations respectively. Visual inspection of CNAs revealed that CNAtra is the most robust detection tool for low-coverage cancer data.

Identification and utilization of copy number information for correcting Hi-C contact map of cancer cell lines

“…The CNV caller module identifies CNVs in cancer cell lines using the RD signal computed by the RD calculator module. It incorporates the multimodal RD signal modeling approach of CNAtra [ 23 ] to hierarchically identify large-scale and focal CN-altered segments. These CN events are integrated to generate the CNV track that is used as an explicit bias source for correcting chromatin interaction matrix.…”

“…Overall results showed that midpoint and exact-cut approaches failed to capture RD signal amplitude change in many cases. Considering the CNVs identified by CNAtra [ 23 ] (ref) using WGS datasets of MCF7 and LNCaP as ground truth, HiCNAtra’s CNV caller outputs using midpoint and exact-cut approaches resulted in several false negatives. For example, in the MCF7 chr 1 (80–100 Mb) region, HiCNAtra’s CNV caller output using the entire-fragment counting method revealed high concordance with the CNAtra’s WGS-derived CNV output of the same cell line.…”

“…HiCNAtra’s CNV caller module adapted the CNAtra approach [ 23 ] to identify CNVs. CNAtra introduced a novel hierarchical CNV detection method based on multimodal distribution to identify both LCVs and FAs in cancer cell lines regardless of the complexity of structural variations.…”

“…HiNT utilizes a generalized additive model (GAM) based on unimodal Poisson distribution, introduced earlier in BIC-seq2 [ 22 ], to normalize the read counts. This leaves plenty of room for improvement for identifying CNVs from Hi-C/3C-seq reads by using a multimodal RD signal distribution which is particularly suited for cancer cell lines [ 23 ].…”

“…RD signal computed by HiCNAtra from Hi-C/3C-seq data successfully recapitulates the WGS-derived RD signal of the corresponding cell line. Our RD-computing approach can extract the RD signal at high resolution that allows precise detection of both large-scale and focal CN alterations using the multimodal distribution-based hierarchical CNV calling approach of CNAtra [ 23 ]. Benchmarking of HiCNAtra’s CNV caller module against OneD and HiNT showed that HiCNAtra is the only tool to accurately estimate the CN profiles of karyotypically abnormal cell lines.…”