Performance of copy number variants detection based on whole-genome sequencing by DNBSEQ platforms

• Main Authors: Junhua Rao, Lihua Peng, Xinming Liang, Hui Jiang, Chunyu Geng, Xia Zhao, Xin Liu, Guangyi Fan, Fang Chen, Feng Mu
• Format: Article
• Language: English
• Published: BMC 2020-11-01
• Series: BMC Bioinformatics
• Subjects: Copy number variant (CNV); Whole-genome sequencing (WGS); DNBSEQ; Benchmark
• Online Access: http://link.springer.com/article/10.1186/s12859-020-03859-x

Abstract Background DNBSEQ™ platforms are new massively parallel sequencing (MPS) platforms that use DNA nanoball technology. Use of data generated from DNBSEQ™ platforms to detect single nucleotide variants (SNVs) and small insertions and deletions (indels) has proven to be quite effective, while the feasibility of copy number variants (CNVs) detection is unclear. Results Here, we first benchmarked different CNV detection tools based on Illumina whole-genome sequencing (WGS) data of NA12878 and then assessed these tools in CNV detection based on DNBSEQ™ sequencing data from the same sample. When the same tool was used, the CNVs detected based on DNBSEQ™ and Illumina data were similar in quantity, length and distribution, while great differences existed within results from different tools and even based on data from a single platform. We further estimated the CNV detection power based on available CNV benchmarks of NA12878 and found similar precision and sensitivity between the DNBSEQ™ and Illumina platforms. We also found higher precision of CNVs shorter than 1 kbp based on DNBSEQ™ platforms than those based on Illumina platforms by using Pindel, DELLY and LUMPY. We carefully compared these two available benchmarks and found a large proportion of specific CNVs between them. Thus, we constructed a more complete CNV benchmark of NA12878 containing 3512 CNV regions. Conclusions We assessed and benchmarked CNV detections based on WGS with DNBSEQ™ platforms and provide guidelines for future studies.