A hybrid computational strategy to address WGS variant analysis in >5000 samples

Abstract

BackgroundThe decreasing costs of sequencing are driving the need for cost effective and real time variant calling of whole genome sequencing data. The scale of these projects are far beyond the capacity of typical computing resources available with most research labs. Other infrastructures like the cloud AWS environment and supercomputers also have limitations due to which large scale joint variant calling becomes infeasible, and infrastructure specific variant calling strategies either fail to scale up to large datasets or abandon joint calling strategies.ResultsWe present a high throughput framework including multiple variant callers for single nucleotide variant (SNV) calling, which leverages hybrid computing infrastructure consisting of cloud AWS, supercomputers and local high performance computing infrastructures. We present a novel binning approach for large scale joint variant calling and imputation which can scale up to over 10,000 samples while producing SNV callsets with high sensitivity and specificity. As a proof of principle, we present results of analysis on Cohorts for Heart And Aging Research in Genomic Epidemiology (CHARGE) WGS freeze 3 dataset in which joint calling, imputation and phasing of over 5300 whole genome samples was produced in under 6 weeks using four state-of-the-art callers. The callers used were SNPTools, GATK-HaplotypeCaller, GATK-UnifiedGenotyper and GotCloud. We used Amazon AWS, a 4000-core in-house cluster at Baylor College of Medicine, IBM power PC Blue BioU at Rice and Rhea at Oak Ridge National Laboratory (ORNL) for the computation. AWS was used for joint calling of 180 TB of BAM files, and ORNL and Rice supercomputers were used for the imputation and phasing step. All other steps were carried out on the local compute cluster. The entire operation used 5.2 million core hours and only transferred a total of 6 TB of data across the platforms.ConclusionsEven with increasing sizes of whole genome datasets, ensemble joint calling of SNVs for low coverage data can be accomplished in a scalable, cost effective and fast manner by using heterogeneous computing platforms without compromising on the quality of variants.Electronic supplementary materialThe online version of this article (doi:10.1186/s12859-016-1211-6) contains supplementary material, which is available to authorized users.

Sequencing of 53,831 diverse genomes from the NHLBI TOPMed Program

“…Here we describe WGS from the first 53,831 TOPMed samples selected from data sets that are now available via dbGaP controlled-access (Supplementary Tables 1 and 2); additional data will be made available as quality control (QC), variant calling and dbGaP curation are completed. Our work identifies and characterizes the rare variants that comprise the majority of human genomic variation 7, [12][13][14] and extends previous efforts that relied on genotyping arrays [15][16][17] , low-coverage WGS 7,8 , exome sequencing 2,12,18 , or analyses of smaller sample collections [19][20][21][22] . Since rare variants represent more recent and potentially more deleterious changes that can impact protein function, gene expression, or other biologically important elements, their discovery and study are crucial for understanding the genetics and biology of human health and disease 13,23,24 .…”

“…We also evaluated potential benefits from high coverage WGS relative to exome sequencing (depth >30X) 18 and low coverage WGS (depth >6X) 22 in 430 Framingham Heart Study samples. TOPMed WGS identified 23.8 million variants in these samples, compared with 20.5 million variants in low coverage sequencing analysis (a 16% increase).…”

Hardy Weinberg Exact Test In Large Scale Variant Calling Quality Control

“…The alignment was done using BWA [16] integrated in the Mercury pipeline [17]. We called biallelic SNPs across all 5297 samples, taking an ensemble variant calling approach goSNAP [18], which employs four variant callers GATK-HaplotypeCaller [19, 20] with gVCF option, GATK-UnifiedGenotyper [19, 20], SNPTools [21] and GotCloud [22], each enforced in a joint calling mode. To ensure a high quality variant call set, we applied a consensus filtering and selected 72,945,834 variant sites which were called at least in 3 out of all 4 callers.…”

Exploring the zoonotic potential of Mycobacterium bovis using variant calling approaches

“…Repositories of SNP data for TB community has been generated and is on the rise [24][25][26][27][28][29], but, all predictions till date have been done using individual samples, which have their own limitations and are known to include false-positives, due to low coverage, small read lengths and sequencing errors [30]. An effort to enlist the variation pro le present within a cohort is still lacking as it becomes computationally challenging to predict SNP/Indel present across samples within a cohort during multi-sample variant prediction [31].…”

“…Approaches like Joint Variant Calling (JVC), (concept used for the rst time on prokaryotes in this study) which predict variants present in a cohort, promises to overcome the shortcomings proposed by single sample variant calling (SVC) methods, as variants are analyzed simultaneously across all samples in a population [31,37,38]. JVC can predict variants for low coverage data in cohorts with high sensitivity.…”