Biology subjects

Asmann, Y.

Publications and source records attributed to Asmann, Y..

2 recordsLinked to original sources

bioRxiv2019

Managing genomic variant calling workflows with Swift/T

Genomic variant discovery is frequently performed using the GATK Best Practices variant calling pipeline, a complex workflow with multiple steps, fans/merges, and conditionals. This complexity makes management of the workflow difficult on a computer cluster, especially when running in parallel on large batches of data: hundreds or thousands of samples at a time. Here we describe a wrapper for the GATK-based variant calling workflow using the Swift/T parallel scripting language. Standard built-in features include the flexibility to split by chromosome before variant calling, optionally permitting the analysis to continue when faulty samples are detected, and allowing users to analyze multiple samples in parallel within each cluster node. The use of Swift/T conveys two key advantages: (1) Thanks to the embedded ability of Swift/T to transparently operate in multiple cluster scheduling environments (PBS Torque, SLURM, Cray aprun environment, etc.,) a single workflow is trivially portable across numerous clusters; (2) The leaf functions of Swift/T permit developers to easily swap executables in and out of the workflow, conditional on the analysts choice, which makes the workflow easy to maintain. This modular design permits separation of the workflow into multiple stages and the request of resources optimal for each stage of the pipeline. While Swift/Ts implicit data-level parallelism eliminates the need for the developer to code parallel analysis of multiple samples, it does make debugging of the workflow a bit more difficult, as is the case with any implicitly parallel code. With the above features, users have a powerful and portable way to scale up their variant calling analysis to run in many traditional computer cluster architectures.\n\nhttps://github.com/ncsa/Swift-T-Variant-Calling\n\nhttp://swift-t-variant-calling.readthedocs.io/en/latest/

bioinformatics↗

bioRxiv2019

Systematic analysis of dark and camouflaged genes: disease-relevant genes hiding in plain sight

BackgroundThe human genome contains dark gene regions that cannot be adequately assembled or aligned using standard short-read sequencing technologies, preventing researchers from identifying mutations within these gene regions that may be relevant to human disease. Here, we identify regions that are dark by depth (few mappable reads) and others that are camouflaged (ambiguous alignment), and we assess how well long-read technologies resolve these regions. We further present an algorithm to resolve most camouflaged regions (including in short-read data) and apply it to the Alzheimers Disease Sequencing Project (ADSP; 13142 samples), as a proof of principle.\n\nResultsBased on standard whole-genome lllumina sequencing data, we identified 37873 dark regions in 5857 gene bodies (3635 protein-coding) from pathways important to human health, development, and reproduction. Of the 5857 gene bodies, 494 (8.4%) were 100% dark (142 protein-coding) and 2046 (34.9%) were [≥]5% dark (628 protein-coding). Exactly 2757 dark regions were in protein-coding exons (CDS) across 744 genes. Long-read sequencing technologies from 10x Genomics, PacBio, and Oxford Nanopore Technologies reduced dark CDS regions to approximately 45.1%, 33.3%, and 18.2% respectively. Applying our algorithm to the ADSP, we rescued 4622 exonic variants from 501 camouflaged genes, including a rare, ten-nucleotide frameshift deletion in CR1, a top Alzheimers disease gene, found in only five ADSP cases and zero controls.\n\nConclusionsWhile we could not formally assess the CR1 frameshift mutation in Alzheimers disease (insufficient sample-size), we believe it merits investigating in a larger cohort. There remain thousands of potentially important genomic regions overlooked by short-read sequencing that are largely resolved by long-read technologies.

genomics↗