agr_jbrowse_vcf

Tools for processing VCF data for JBrowse

This docker file is a collection of tools for fetching Alliance VCF files, processing them into tabix indexed files, and then moving the results to the Alliance S3 bucket (agrjbrowse by default, though agrjbrowse2 has been used for testing in the past). The use of this docker file is controlled in GoCD in the JBrowseSoftwareProcessVCF and JBrowseProcessVCF pipelines. The primary description of the workflow for getting genome browsing ready for a new release is described in the agr_jbrowse_gff repo (https://github.com/alliance-genome/agr_jbrowse_gff).

Typical workflow

1. After both the phenotypic variant VCF and high throughput VCF files are available in the FMS, edit the parallel.sh file to update the value on the RELEASE=... line. Commit and push that change.

2. Unpause the JBrowseSoftwareProcessVCF and JBrowseProcessVCF pipelines in GoCD (generally they are kept paused to avoid accidentally running them).

Important note about VCF updates in the FMS

It is typical for the "latest" VCFs to be updated for a given release very near the release, as (I think) VEP is run "on last time". As a result, the JBrowseProcessVCF pipeline will have to be run after that update, and the Apollo build pipeline, ApolloSoftwareStage, will have to be run after that to make sure the latest VCFs are in the Apollo container.