Genome Nexus Annotation Pipeline

Annotation of genomic variants from a MAF for import into the cBioPortal using Genome Nexus.

MAF Annotation

The annotationPipeline module is a command line tool to annotate a maf using genome nexus.

Build From Source

1. Create application.properties file:

   cp annotationPipeline/src/main/resources/application.properties.EXAMPLE annotationPipeline/src/main/resources/application.properties

   [!TIP] If you have your own installation of Genome Nexus, you can point to it by modifying the genomenexus.base=<URL> in application.properties

2. Create log4j.properties:

cp annotationPipeline/src/main/resources/log4j.properties.console.EXAMPLE annotationPipeline/src/main/resources/log4j.properties

Build

mvn clean install

Run Annotate

To use it, build the project using maven and run it like so:

java -jar annotationPipeline/target/annotationPipeline-*.jar \
    --filename <INPUT_MAF> \
    --output-filename <OUTPUT DESTINATION> \
    --isoform-override <mskcc or uniprot>

To output error reporting to a file, supply the -e option a location for the file to be saved. By running the jar without any arguments or by providing the optional parameter -h you can view the full usage statement.

Annotate with Docker

Genome Nexus Annotation Pipeline is available on DockerHub: https://hub.docker.com/r/genomenexus/gn-annotation-pipeline.

Usage instruction

docker pull genomenexus/gn-annotation-pipeline:master 

docker run -v ${PWD}:/wd genomenexus/gn-annotation-pipeline:master java -jar annotationPipeline.jar --filename /wd/input.txt  --output-filename /wd/output.txt

• -v ${PWD}:/wd: This option maps the current working directory to a volume within the Docker container at the path /wd. This makes it possible for files in the host directory to be accessed and modified between container and host.
• --filename /wd/input.txt: This option specifies the input file location at /wd/input.txt, which should be under the same directory as output file.
• --output-filename /wd/output.txt: This option specifies the output file where the annotated results will be saved. The file will be created at /wd/output.txt, which should be under the same directory as input file.

Logging

[!IMPORTANT] Logging via docker has been changed to stdout by default since v1.0.4

Optional parameters

Reference Genome

The Genome Nexus Annotation Pipeline supports two versions of the human genome reference assembly:

1. GRCh37 (default)
2. GRCh38

[!NOTE] By default, the pipeline uses GRCh37.

Using GRCh38

If you want to annotate with GRCh38, please set the GENOMENEXUS_BASE environment variable to https://grch38.genomenexus.org. Here's an example of how to do this:

docker run -e GENOMENEXUS_BASE=https://grch38.genomenexus.org -v ${PWD}:/wd genomenexus/gn-annotation-pipeline:master --filename /wd/input.txt --output-filename /wd/output.txt --isoform-override uniprot

Annotation fields

Add additional annotation columns

Genome Nexus supports additional annotation columns with the setting of "enrichment_fields". The configuration for these enrichment fields is managed through the application.properties file, please refer to Pre-build steps section.

To configure the enrichment fields, you need to include the desired field names from the provided list in the -Dgenomenexus.enrichment_fields= parameter of the command line, or directly add field names in genomenexus.enrichment_fields= in application.properties file. Multiple field names can be specified by separating them with commas. annotation_summary is highly recommended to add as default since it's crucial for lots of annotation fields.

Example:

java -Dgenomenexus.enrichment_fields=annotation_summary,my_variant_info \
    -jar annotationPipeline/target/annotationPipeline-*.jar \ 
    -r \ 
    --filename test/data/minimal_example.in.txt \ 
    --output-filename test/data/minimal_example.out.uniprot.txt \ 
    --isoform-override uniprot

Available enrichment fields:

• annotation_summary:
  • This field is essential for most annotation processes.
• my_variant_info
  • This field provides gnomAD information. Genome Nexus caches static myvariant.info annotation into database. Upon adding my_variant_info to the enrichment_fields, Genome Nexus will query the existing variant information from the myvariant.info cache within the database for relevant variants. For new variants, Genome Nexus will retrieve the data from myvariant.info and store it in the database.
• polyphen
• sift
• mutation_assessor
  • 'mutation_assessor' provides V3 version annotation from Mutation Assessor, which is currently stored in Genome Nexus database
• nucleotide_context
• oncokb:
  • 'oncokb' provides annotations of the biological consequences and clinical implications from OncoKB website. OncoKB token is required (see more information from: https://www.oncokb.org/apiAccess). Please also provide your token in -Doncokb.token=abc123 command line parameter, or directly add oncokb.token=abc123in the application.properties. No OncoKB annotation columns will be added if no valid token is provided

Minimal MAF Example

For an example minimal input file see test/data/minimal_example.in.txt and corresponding output test/data/minimal_example.out.uniprot.txt. The output file was generated with:

$JAVA_HOME/bin/java -jar annotationPipeline/target/annotationPipeline-*.jar \
    -r \
    --filename test/data/minimal_example.in.txt  \
    --output-filename test/data/minimal_example.out.uniprot.txt \
    --isoform-override uniprot

Direct Database Annotation

There used to be a utility/module called databaseAnnotator which could be used to annotate database records which were already loaded into a cBioPortal database. This module was not being maintained and has been removed (code is still reachable through git history).

Annotator

The annotator module is the client code that makes calls to the Genome Nexus server and interprets the response. The annotationPipeline module uses this as a dependency.

Updating the Genome Nexus Annotation Pipeline

The annotation pipeline uses models brought in by the auto-generated genome-nexus-java-api-client here and here .

In the event of an update to the Genome Nexus server (e.g adding to an existing model), the genome-nexus-java-api-client must be regenerated and propagated to this codebase. To generate the newest genome-nexus-java-api-client, refer to instructions here.

Once updated, update the pom with the newest commit hash from the genome-nexus-java-api-client codebase.