Informatics Tools for High-Throughput Sequences Data Analysis

Broad Institute, Inc., Cambridge MA

Investigators

Linked publications & trials

Abstract

DESCRIPTION (provided by applicant): The Genome Analysis Toolkit (GATK) is a suite of best-in-class, widely-used, well-supported, open-source tools for processing and analysis of next-generation DNA sequencing (NGS) data. These tools currently include a multiple sequence realigner, a covariate-correcting base quality score recalibrator, multi-sample SNP, INDEL, and CNV genotypers, machine learning algorithms for false positive identification, variant evaluation modules, somatic SNP and indel callers, and hundreds of other tools. Underlying all of these tools is our structured programming framework (GATK-Engine) that uses the functional programming philosophy of MapReduce to make writing feature-rich, efficient and robust analysis tools easy. By centralizing common data management infrastructure, all GATK-based tools benefit from the engine's correctness, CPU and memory efficiency, as well as automatic distributed and shared memory parallelization, essential capabilities given the massive and growing size of NGS datasets. The GATK currently supports all of the major sequencing technologies including lllumina. Life Sciences 454, and ABI SOLID, from hybrid capture of exomes to 1000s of low-pass samples in the 1000 Genomes Project. Our emphasis on technology-agnostic processing tools has helped to popularize the now standard SAM/BAM and VCFs formats for representing NGS data and variation calls, respectively. In this RFA we propose to continue to develop the GATK-Engine and data processing tools to (1) achieve complete and accurate variation discovery and genotyping for all major sequencing study designs and NGS technologies (2) optimize the GATK-Engine and pipelining infrastructure to operate efficiently on distributed data sets at the scale of tens of thousands of samples (3) extend the GATK data processing tools to support the upcoming sequencing technologies of Complete Genomics, lon Torrent, and Pacific Biosciences as well as we do current technologies, (4) expand significantly our educational and support structures to ensure that the longtail of future NGS users can benefit from the best-practice data processing and analysis tools in the GATK.