github.com/DataBiosphere/xvcfmerge/xvcfmerge

Fuente: Zenodo
Salvato in:
Dettagli Bibliografici
Autori principali: Brian Hannafious, Lon Blauvelt
Natura: Recurso digital
Pubblicazione: Zenodo 2022
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866901690719928320
author Brian Hannafious
Lon Blauvelt
author_facet Brian Hannafious
Lon Blauvelt
contents <h1>xvcfmerge</h1> <p>Define a WDL workflow to merge jointly called VCF files. This workflow accepts input as array of reference URIs. The reference URIs should be either Google Storage URLs or <a href="https://support.terra.bio/hc/en-us/articles/360039330211-Data-Access-with-the-GA4GH-Data-Repository-Service-DRS-">DRS</a> URIs.</p> <h2>Description</h2> <p><a href="https://github.com/DataBiosphere/xsamtools">xsamtools</a> is used to perform the merge. Each input VCF should describe a single chromosome and should have equivalent headers.</p> <p>In order to keep hard disk requirements small and improve performance, input/output objects are streamed directly from cloud storage using FIFO queues (named pipes).</p> <p>An appropriate workflow configuration is ~16GB RAM, ~8 CPUs, and standard storage.</p> <h2>Runtime</h2> <p>Runtime varies with the number of VCFs to merge, the number of samples, runtime configuration, and other factors. The table below may be helpful to understand the expected runtime for your workflow.</p> <p>| cpu | memory | number of vcfs | average size | samples | runtime | | --- | ------ | -------------- | ------------ | ------- | --------- | | 2 | 2GB | 2 | 2GB | 2 | 0.9 hours | | 8 | 64GB | 2 | 2GB | 2 | 0.7 hours | | 8 | 64GB | 2 | 5GB | 5370 | 4 hours | | 8 | 64GB | 5 | 5GB | 4100 | 18 hours |</p> <h2>Tests</h2> <p>Test workflows are executed locally with <a href="https://github.com/chanzuckerberg/miniwdl">miniwdl</a>, and require access to the <a href="https://terra.biodatacatalyst.nhlbi.nih.gov/#workspaces/firecloud-cgl/terra-notebook-utils-tests">terra-notebook-utils-tests</a> Terra workspace and GCP authorization.</p> <p>To run basic tests,</p> <pre><code>docker login gcloud auth login make test </code></pre> <p>Note that the WDL built-in function size() does not function in miniwdl, which is used in the makefile, when working with gs URIs which may lead to a failing test unless lines involving that function are commented out. size() functions correctly on Cromwell and is a necessary inclusion for GCP-based backends including Terra.</p> <p>More robust tests against DRS resolved objects should be executed prior to release with</p> <pre><code>gcloud auth application-default login make test_drs </code></pre> <p>Please contact Lon (lblauvel at ucsc dot edu) for data and workspace access issues.</p>
format Recurso digital
id zenodo_https___doi_org_10_5281_zenodo_15610743
institution Zenodo
language
publishDate 2022
publisher Zenodo
record_format zenodo
spellingShingle github.com/DataBiosphere/xvcfmerge/xvcfmerge
Brian Hannafious
Lon Blauvelt
utilities
xsamtools
xvcf
<h1>xvcfmerge</h1> <p>Define a WDL workflow to merge jointly called VCF files. This workflow accepts input as array of reference URIs. The reference URIs should be either Google Storage URLs or <a href="https://support.terra.bio/hc/en-us/articles/360039330211-Data-Access-with-the-GA4GH-Data-Repository-Service-DRS-">DRS</a> URIs.</p> <h2>Description</h2> <p><a href="https://github.com/DataBiosphere/xsamtools">xsamtools</a> is used to perform the merge. Each input VCF should describe a single chromosome and should have equivalent headers.</p> <p>In order to keep hard disk requirements small and improve performance, input/output objects are streamed directly from cloud storage using FIFO queues (named pipes).</p> <p>An appropriate workflow configuration is ~16GB RAM, ~8 CPUs, and standard storage.</p> <h2>Runtime</h2> <p>Runtime varies with the number of VCFs to merge, the number of samples, runtime configuration, and other factors. The table below may be helpful to understand the expected runtime for your workflow.</p> <p>| cpu | memory | number of vcfs | average size | samples | runtime | | --- | ------ | -------------- | ------------ | ------- | --------- | | 2 | 2GB | 2 | 2GB | 2 | 0.9 hours | | 8 | 64GB | 2 | 2GB | 2 | 0.7 hours | | 8 | 64GB | 2 | 5GB | 5370 | 4 hours | | 8 | 64GB | 5 | 5GB | 4100 | 18 hours |</p> <h2>Tests</h2> <p>Test workflows are executed locally with <a href="https://github.com/chanzuckerberg/miniwdl">miniwdl</a>, and require access to the <a href="https://terra.biodatacatalyst.nhlbi.nih.gov/#workspaces/firecloud-cgl/terra-notebook-utils-tests">terra-notebook-utils-tests</a> Terra workspace and GCP authorization.</p> <p>To run basic tests,</p> <pre><code>docker login gcloud auth login make test </code></pre> <p>Note that the WDL built-in function size() does not function in miniwdl, which is used in the makefile, when working with gs URIs which may lead to a failing test unless lines involving that function are commented out. size() functions correctly on Cromwell and is a necessary inclusion for GCP-based backends including Terra.</p> <p>More robust tests against DRS resolved objects should be executed prior to release with</p> <pre><code>gcloud auth application-default login make test_drs </code></pre> <p>Please contact Lon (lblauvel at ucsc dot edu) for data and workspace access issues.</p>
title github.com/DataBiosphere/xvcfmerge/xvcfmerge
topic utilities
xsamtools
xvcf
url https://doi.org/10.5281/zenodo.15610743