Skip to content

Latest commit

 

History

History
executable file
·
103 lines (83 loc) · 3.15 KB

README.md

File metadata and controls

executable file
·
103 lines (83 loc) · 3.15 KB
Raw

WARNING: Not actively maintained!

Build the annotator

This tutorial builds the tools needed to annotate GRCh37 or GRCh38 of the human genome reference.

Container Builder and dsub are used to run all of these steps in the cloud.

(1) Configure project variables.

Set a few environment variables to facilitate cutting and pasting the subsequent commands.

# The Google Cloud Platform project id in which the Docker containers
# will be built and stored.
PROJECT_ID=your-project-id
# The bucket name (with the gs:// prefix) where the cached version of
# dbNSFP should be stored.
BUCKET=gs://your-bucket-name

(2) Build the VEP Docker container.

Run one of the commands below to build a Docker container that is configured to run VEP on human genetic variants in GRCh37 or GRCh38 coordinates with annotations including dbNSFP, SIFT, and many others. Both of these commands can be run in parallel if you wish to annotate using both reference genomes.

GRCh37

gcloud container builds submit \
    --substitutions=_GENOME_ASSEMBLY=GRCh37,_ENSEMBL_RELEASE=89,_DBNSFP_BASE=${BUCKET}/dbNSFPv2.9.3/dbNSFP,_CONTAINER_SUFFIX=_89_grch37:latest \
    --config=vep_container.yaml \
    .

GRCh38

gcloud container builds submit \
    --substitutions=_GENOME_ASSEMBLY=GRCh38,_ENSEMBL_RELEASE=89,_DBNSFP_BASE=${BUCKET}/dbNSFPv3.4c/dbNSFP,_CONTAINER_SUFFIX=_89_grch38:latest \
    --config=vep_container.yaml \
    .

(3) Cache dbNSFP.

First run the command below to create the Docker container with tools and scripts needed to process dbNSFP. This Docker container can be used for any version of dbNSFP.

gcloud --project ${PROJECT_ID} container builds submit \
    --substitutions=_CONTAINER_TAG=:latest \
    --config=dbNSFP_container.yaml \
    .

Then run the container via dsub to download and cache dbNSFP annotations in Cloud Storage. These commands can be run in parallel if you wish to annotate using both reference genomes.

  • Note that it can take several hours for the job to complete.
  • The values for FILEID in the commands came from dbNSFP documentation where you can also get detail on other available versions of dbNSFP.

GRCh37

dsub \
  --project ${PROJECT_ID} \
  --image gcr.io/${PROJECT_ID}/dbnsfp_cache_builder:latest \
  --zones "us-central1-*" \
  --disk-size 200 \
  --min-cores 8 \
  --logging ${BUCKET}/dbNSFPv3.4c/dbNSFPv3.4c.log \
  --env FILEID=0B60wROKy6OqcaWJ4Y0xvR2k1aUU \
  --output-recursive OUTPUT_PATH=${BUCKET}/dbNSFPv3.4c/ \
  --command '/opt/download_dbNSFP.sh &&
             /opt/build_databases.sh'

GRCh38

dsub \
  --project ${PROJECT_ID} \
  --image gcr.io/${PROJECT_ID}/dbnsfp_cache_builder:latest \
  --zones "us-central1-*" \
  --disk-size 200 \
  --min-cores 8 \
  --logging ${BUCKET}/dbNSFPv2.9.3/dbNSFPv2.9.3.log \
  --env FILEID=0B60wROKy6OqceTNZRkZnaERWREk \
  --output-recursive OUTPUT_PATH=${BUCKET}/dbNSFPv2.9.3/ \
  --command '/opt/download_dbNSFP.sh &&
             /opt/build_databases.sh'