Files in this item



application/pdfECE499-Sp2016-li-Huiren.pdf (887kB)Restricted to U of Illinois
(no description provided)PDF


Title:Exploration of GPU acceleration of K-mer counting (KMC2) and error correction tool (BLESS)
Author(s):Li, Huiren
Contributor(s):Chen, Deming
Degree:B.S. (bachelor's)
Subject(s):k-mer counting
Bloom-filter-based error correction solution
Abstract:Next-generation sequencing (NGS) technology has led to a rapid growth in the amount of genomic information together with more errors in the reads. An algorithm named Bloom-filter-based Error Correction Solution (BLESS) developed by Professor Deming Chen’s group at the University of Illinois is a memory frugal error correction tool that also demonstrated the best error correction quality for NGS reads. One major concern about BLESS is the running time. In order to reduce the running time of BLESS, previous work has been done to move the calculation to the GPU side. By modifying the existing GPU algorithm, we have achieved a speedup from 9x to 12x on a commercial GPU, GTX 770. Another time-consuming task in error correction is k-mer counting in the fastq files. KMC2 is a popular k-mer counting tool developed by Sebastian Deorowicz. By exporting the distribution stage to GPU, we achieved an average of 2.65x speedup for this stage.
Issue Date:2016-05
Date Available in IDEALS:2016-08-29

This item appears in the following Collection(s)

Item Statistics