Removing duplicate reads using graphics processing units

Abstract

BackgroundDuring library construction polymerase chain reaction is used to enrich the DNA before sequencing. Typically, this process generates duplicate read sequences. Removal of these artifacts is mandatory, as they can affect the correct interpretation of data in several analyses. Ideally, duplicate reads should be characterized by identical nucleotide sequences. However, due to sequencing errors, duplicates may also be nearly-identical. Removing nearly-identical duplicates can result in a notable computational effort. To deal with this challenge, we recently proposed a GPU method aimed at removing identical and nearly-identical duplicates generated with an Illumina platform.The method implements an approach based on prefix-suffix comparison. Read sequences with identical prefix are considered potential duplicates. Then, their suffixes are compared to identify and remove those that are actually duplicated.Although the method can be efficiently used to remove duplicates, there are some limitations that need to be overcome. In particular, it cannot to detect potential duplicates in the event that prefixes are longer than 27 bases, and it does not provide support for paired-end read libraries. Moreover, large clusters of potential duplicates are split into smaller with the aim to guarantees a reasonable computing time. This heuristic may affect the accuracy of the analysis.ResultsIn this work we propose GPU-DupRemoval, a new implementation of our method able to (i) cluster reads without constraints on the maximum length of the prefixes, (ii) support both single- and paired-end read libraries, and (iii) analyze large clusters of potential duplicates.ConclusionsDue to the massive parallelization obtained by exploiting graphics cards, GPU-DupRemoval removes duplicate reads faster than other cutting-edge solutions, while outperforming most of them in terms of amount of duplicates reads.

Transcriptomic Analysis of Rhodococcus opacus R7 Grown on o-Xylene by RNA-Seq

“…Quality of raw sequencing data was checked with FastQC (version 0.11.5). Then, sequencing reads were trimmed using Trim Galore (version 0.4.4) (Krueger, 2015), while duplicates were identified and removed using GPU-Dup Removal (Manconi et al, 2016). After data pre-processing FastQC reported an acceptable level of quality with a median over 30 and a vast majority of 100 nt-long sequences.…”

Transcriptomic analysis of Rhodococcus opacus R7 grown on polyethylene by RNA-seq

“…The RNA sequencing of the samples was performed by Illumina HiSeq platform (Biodiversa, Italy) generating about 110 million and 113 million pairs of raw reads (forward and reverse strands) for the three malate samples and the three PE samples, respectively. The quality of raw sequencing data was checked with Trim Galore (version 0.4.4) (URL: http://www.bioinformatics.babraham.ac.uk/projects/trim_galore ) to trim the sequencing reads 26 , while GPU-Dup Removal was applied to identify and remove duplicates 27 . The quality control resulted in about 92 million and 60 million reads for malate and PE samples, respectively.…”

Fast-HBR: Fast hash based duplicate read remover

“…One of the preprocessing steps that reduce the dataset size is removing duplicate reads in the dataset. This step is essential for sequence-based algorithms since duplicate reads affect the algorithm accuracy [ 4 ]. Removing duplicate reads may reduce the assembly algorithms consumption of RAM [ 5 ].…”

“…essential for sequence-based algorithms since duplicate reads affect the algorithm accuracy [4]. Removing duplicate reads may reduce the assembly algorithms consumption of RAM [5].…”