Abstract
<p class="Els-1storder-head">Genomics and Next Generation Sequencers (NGS) like Illumina Hiseq produce data in the order of 200 billion base pairs in a single one-week run for a 60x human genome coverage, which requires modern high-throughput experimental technologies that can only be tackled with high performance computing (HPC) and specialized software algorithms called “short read aligners”. This paper focuses on the implementation of the DNA sequencing as a set of MapReduce programs that will accept a DNA data set as a FASTQ file and finally generate a VCF (variant call format) file, which has variants for a given DNA data set. In this paper MapReduce/Hadoop along with Burrows-Wheeler Aligner (BWA), Sequence Alignment/Map (SAM) tools, are fully utilized to provide various utilities for manipulating alignments, including sorting, merging, indexing, and generating alignments. The Map-Sort-Reduce process is designed to be suited for a Hadoop framework in which each cluster is a traditional N-node Hadoop cluster to utilize all of the Hadoop features like HDFS, program management and fault tolerance. The Map step performs multiple instances of the short read alignment algorithm (BoWTie) that run in parallel in Hadoop. The ordered list of the sequence reads are used as input tuples and the output tuples are the alignments of the short reads. In the Reduce step many parallel instances of the Short Oligonucleotide Analysis Package for SNP (SOAPsnp) algorithm run in the cluster. Input tuples are sorted alignments for a partition and the output tuples are SNP calls. Results are stored via HDFS, and then archived in SOAPsnp format. The proposed framework enables extremely fast discovering somatic mutations, inferring population genetical parameters, and performing association tests directly based on sequencing data without explicit genotyping or linkage-based imputation. It also demonstrate that this method achieves comparable accuracy to alternative methods for sequencing data processing.</p><p class="Abstract"><em></em><em><br /></em></p>
Highlights
Today, genome sequencing machines are able to generate thousands of gigabases of DNA and RNA sequencing data in a few hours for less than US$1,000 [1, 2].Success in biology and the life sciences depends on our ability to properly analyze the big data sets that are generated by these technologies, which in turn requires us to adopt advances in informatics
According to [3] “if finding DNA was the discovery of the exact substance holding our genetic makeup information, DNA sequencing is the discovery of the process that will allow us to read that information.”
Variant detection is the process of finding bases in the Next Generation Sequencers (NGS) data that differ from the reference genome, such as hg18 or hg19; these refer to the version of the human genome assembly and determine the version of the corresponding reference annotations
Summary
Genome sequencing machines (such as Illumina’s HiSeq 4000) are able to generate thousands of gigabases of DNA and RNA sequencing data in a few hours for less than US$1,000 (a few years ago, the price was over US$100,000, and sequencing the first human genome cost about US$3 billion) [1, 2]. 2. The input data (FASTQ data) size is big (a single DNA sequence sample can be up to 900 GB). 3. With a single powerful server, it takes too long (up to 80 hours) to process one DNA sequence and extract variants such as single nucleotide polymorphisms (SNPs). One important example is the identification of single nucleotide polymorphisms (SNPs). The identification and extraction of SNPs from raw genetic sequences involves many algorithms and the application of a diverse set of tools iJES ‒ Volume 4, Issue 4, 2016
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
