PECC: Correcting contigs based on paired-end read distribution. (August 2017)

• Record Type: Journal Article
• Authors: Li, Min; Wu, Binbin; Yan, Xiaodong; Luo, Junwei; Pan, Yi; Wu, Fang-Xiang; Wang, Jianxin
• Is Part Of: Computational biology and chemistry. Volume 69(2017)
• Page Start: 178
• Page End: 184
• Publication Date: 2017-08
• Subjects: Next generation sequencing -- De novo assembly -- Contigs -- Paired-end reads; Chemistry -- Data processing -- Periodicals; Biology -- Data processing -- Periodicals; Biochemistry -- Data processing; Biology -- Data processing; Molecular biology -- Data processing; Periodicals; Electronic journals; 542.85
• DOI: 10.1016/j.compbiolchem.2017.03.012
• Languages: English
• ISSNs: 1476-9271

Abstract

Abstract: Motivation: Cheap and fast next generation sequencing (NGS) technologies facilitate research of de novo assembly greatly. The reliability of contigs is critical to construct reliable scaffolding. However, contigs generated from most assemblers contain errors because of the limitation of assembly strategy and computation complexity. Among all these errors, the misassembly error is one of the most harmful types. Results: In this paper, we propose a new method named "PECC" to identify and correct misassembly errors in contigs based on the paired-end read distribution. PECC extracts sequence regions with lower paired-end reads supports and verifies them based on the distribution of paired-end supports. To validate the effectiveness of PECC, we applied PECC to the contigs produced by five popular assemblers on four real datasets, and we also carried out experiments to analyze the influences of PECC on scaffolding. The results show that PECC can reduce misassembly errors and improve the performance of scaffolding results, which demonstrate the promising applications of PECC in de novo assembly.