Skip to content

Latest commit

 

History

History
169 lines (130 loc) · 8.86 KB

README.md

File metadata and controls

169 lines (130 loc) · 8.86 KB
Raw

OncoKB Core

Repository for OncoKB, a precision oncology knowledge base.

The core of OncoKB Annotation service.

Status

Application CI Unit Tests Release Management Sentrey Release

Info

Running Environment

Please confirm your running environment is:

  • Java version: 8
  • MySQL version: 5.7.28

Prepare properties files

cp -r core/src/main/resources/properties-EXAMPLE core/src/main/resources/properties

Properties file

  1. database.properties
    • jdbc.driverClassName : We use mysql as database. Here, it will be com.mysql.jdbc.Driver
    • jdbc.url: Database url
    • jdbc.username & jdbc.password: MySQL user name and password
  2. config.properties

Build the WAR file

mvn clean install -P public -DskipTests=true

The WAR file is under /web/target/

Deploy with frontend

Please choose one of the profile when building the war file

  • curate - core + API + curation website
  • public - core + API + public website (deprecated)

You could find specific instructions in curate or public repo,

Run with Docker containers

OncoKB™ is a precision oncology knowledge base developed at Memorial Sloan Kettering Cancer Center that contains biological and clinical information about genomic alterations in cancer. OncoKB uses Genome Nexus to annotate genomic change to protein change using OncoKB picked transcripts. By default, the API requests are sent to www.genomenexus.org for GRCh37 and grch38.genomenexus.org for GRCh38. However, you can choose to use a local version of Genome Nexus by following the instructions for Option A, otherwise follow instructions for Option B.

OncoKB docker compose file consists of the following services:

  • OncoKB: provides variant annotations

  • OncoKB Transcript: serves OncoKB metadata including gene, transcript, sequence, etc.

  • Genome Nexus: provides annotation and interpretation of genetic variants in cancer

    • GRCh37 (optional):
      • gn-spring-boot: the backend service responsible for aggregating variant annotations from various sources
      • gn-mongo: variants fetched from external resources and small static data are cached in the MongoDB database
      • gn-vep: is a spring boot REST wrapper service for VEP using GRCh37 data
    • GRCh38 (optional):
      • gn-spring-boot-grch38: same as gn-spring-boot service, however the VEP URL points to gn-vep-grch38
      • gn-mongo-grch38: contains static data relevant to GRCh38
      • gn-vep-grch38: a spring boot REST wrapper service for VEP using GRCh38 data

Option A: With Local installation of Genome Nexus

For this option, you need to download the VEP cache, which is used in the gn-vep and gn-vep-grch38 services. We have pre-downloaded the VEP data and saved them to our AWS S3 Bucket. If interested, here are the instructions we followed to download the Genome Nexus VEP Cache.

  1. OncoKB requires a MySQL server and the oncokb and oncokb-transcript databases imported. This step must be completed before continuing the installation process. Reach out to contact@oncokb.org to get access to the data dump.

  2. Download the Genome Nexus VEP data from our AWS S3 Bucket.

    # The home directory is used to store the VEP cache in this tutorial, but this can be changed to your preferred download location.
cd ~
mkdir gn-vep-data && cd "$_"

mkdir 98_GRCh37 && cd "$_"
curl -o 98_GRCh37.tar https://oncokb.s3.amazonaws.com/gn-vep-data/98_GRCh37/98_GRCh37.tar
tar xvf 98_GRCh37.tar

cd ..
mkdir 98_GRCh38 && cd "$_"
curl -o 98_GRCh38.tar https://oncokb.s3.amazonaws.com/gn-vep-data/98_GRCh38/98_GRCh38.tar
tar xvf 98_GRCh38.tar

  3. Set environment variable for the location of VEP caches

    # Update path if the VEP data was installed elsewhere
export VEP_CACHE=~/gn-vep-data/98_GRCh37
export VEP_GRCH38_CACHE=~/gn-vep-data/98_GRCh38

  4. Run docker-compose to create containers.

    docker-compose --profile genome-nexus up -d

    Note: The --profile argument is used as a way to selectively enable services. Services with the genome-nexus profile will only be spun up when the profile is specified.

Option B: Without local installation of Genome Nexus

  1. OncoKB requires a MySQL server and the oncokb and oncokb-transcript databases imported. This step must be completed before continuing the installation process. Reach out to contact@oncokb.org to get access to the data dump.
  2. Remove -Dgenome_nexus.grch37.url and -Dgenome_nexus.grch38.url properties from the oncokb service.
  3. Run docker-compose to spin up oncokb and oncokb-transcript services 
    docker-compose up -d

Additional Information

Generating oncokb-transcript token

The docker compose file has a pre-generated oncokb-transcript JWT token, which is required to make API requests to the oncokb-transcript service. To generate the JWT token, go to the https://jwt.io/ website and follow these instructions:

  1. Add the auth key and set it to ROLE_ADMIN to grant roles. The payload section should look something like this: 
    {
    "sub": "1234567890",
    "name": "John Doe",
    "auth":"ROLE_ADMIN",
    "iat": 1516239022
}
  2. In the Verify Signature section, check the box secret base64 encoded. Copy and paste the oncokb-transcript base64 secret into the input box.
    • You can also change the default base64 secret used for encoding by generating a base64 string and add the environment variable, JHIPSTER_SECURITY_AUTHENTICATION_JWT_BASE64_SECRET: <new-base64-string>, to oncokb-transcript.
  3. Replace -Doncokb_transcript.token with the JWT token you generated.

Generating new VEP data

OncoKB predownloads VEP data and saves it to AWS S3 bucket. These steps are for OncoKB developers and show how to download and upload new Ensembl VEP data to S3. However, you can follow along and save VEP data to your own S3 bucket.

  1. Change Ensembl image in genome-nexus-vep Dockerfile to desired version
  2. Follow instructions to download VEP cache files and FASTA files for GRCh37 and GRCh38.
  3. After downloading your directory should like:
VEP_CACHE/
├─ homo_sapiens/
│  ├─ 98_GRCh37/
│  ├─ 98_GRCh38/
  1. Zip the files
tar cf 98_GRCh37.tar homo_sapiens/98_GRCh37
tar cf 98_GRCh38.tar homo_sapiens/98_GRCh38
  1. Go to AWS S3 webpage and under oncokb/gn-vep-data/, create two folders:
98_GRCh37/
98_GRCh38/
  1. Upload tar files to corresponding S3 folders
  2. Make the two S3 folders (oncokb/gn-vep-data/98_GRCh37/ and oncokb/gn-vep-data/98_GRCh38/) publicly accessible
  3. Update gn-vep and gn-vep-grch38 services in docker-compose.yml
Modify environment variable to point to the new FASTA file

gn-vep
VEP_FASTAFILERELATIVEPATH=homo_sapiens/98_GRCh37/Homo_sapiens.GRCh37.75.dna.primary_assembly.fa.gz

gn-vep-grch38
VEP_FASTAFILERELATIVEPATH=homo_sapiens/98_GRCh38/Homo_sapiens.GRCh38.dna.toplevel.fa.gz
  1. Modify Dockerfile line in genome-nexus-vep to use the new Ensembl VEP image. As of 4/28/2023, genome-nexus-vep uses ensemblorg/ensembl-vep:release_98.3.
  2. Push new genome-nexus-vep image to DockerHub
  3. Change the image for both gn-vep and gn-vep-grch38 to the image built in step 7.

Questions?

The best way is to send an email to contact@oncokb.org so all our team members can help.