Correction of sequencing errors in a mixed set of reads

Open Access
Publisher Website
Google Scholar
Abstract
Motivation: High-throughput sequencing technologies produce large sets of short reads that may contain errors. These sequencing errors make de novo assembly challenging. Error correction aims to reduce the error rate prior assembly. Many de novo sequencing projects use reads from several sequencing technologies to get the benefits of all used technologies and to alleviate their shortcomings. However, combining such a mixed set of reads is problematic as many tools are specific to one sequencing platform. The SOLiD sequencing platform is especially problematic in this regard because of the two base color coding of the reads. Therefore, new tools for working with mixed read sets are needed. Results: We present an error correction tool for correcting substitutions, insertions and deletions in a mixed set of reads produced by various sequencing platforms. We first develop a method for correcting reads from any sequencing technology producing base space reads such as the SOLEXA/Illumina and Roche/454 Life Sciences sequencing platforms. We then further refine the algorithm to correct the color space reads from the Applied Biosystems SOLiD sequencing platform together with normal base space reads. Our new tool is based on the SHREC program that is aimed at correcting SOLEXA/Illumina reads. Our experiments show that we can detect errors with 99% sensitivity and >98% specificity if the combined sequencing coverage of the sets is at least 12. We also show that the error rate of the reads is greatly reduced. Availability: The JAVA source code is freely available at http://www.cs.helsinki.fi/u/lmsalmel/hybrid-shrec/ Contact:leena.salmela@cs.helsinki.fi

This publication has 13 references indexed in Scilit: