SERVE harmonizes various sources of evidence into a single unified model that can be readily used to interpret genomic analyses:
- A model is generated which allows mapping of genomic events to clinical evidence.
- An overview of mutations that are implied to be potential cancer drivers is generated.
In addition, this repo provides a number of utility applications to ingest and analyze various SERVE input sources.
| Module | Description |
|---|---|
| Datamodel | The datamodel that other tools can reuse. |
| Algo | The actual SERVE algorithm. |
| CKB-Importer | Importer of CKB FLEX datamodel. |
| iClusion-Importer | Importer of iClusion datamodel. |
| VICC-Importer | Importer of VICC datamodel. |
To release a new version of the serve and all submodules, perform the following:
git tag ${new_version}
git push origin ${new_version}This will automatically trigger a cloud build instance which will deploy the artifacts to both artifact registry and container registry.
Note the new version should be of the format major.minor.patch where:
- Major indicates a non-backward compatible change (avoid these if possible!)
- Minor indicates a new feature
- Patch indicates a bug fix
Currently, the GitHub release is not automatically created, so you need to create a new release on the GitHub website and attach the respective jar and database generation scripts as additional resources. The jars can be built by running these commands:
mvn versions:set -DnewVersion=${new_version}mvn clean packagemvn versions:revert
In addition, when creating releases on the GitHub website it is convenient to create additional tags with the module prefix (
e.g. serve-v${new_version} when releasing algo)