gms | German Medical Science

Introduction: The annotation of genetic variants is an important step in high-throughput sequencing data analysis pipelines. In the context of cancer sequencing, the annotation of somatic variants plays an essential role for clinical interpretation. As part of the BMBF-funded project MIRACUM, an annotation pipeline for somatic variants (MIRACUM-Pipe) was designed and implemented. One major focus of the pipeline is to link genetic variants to possible treatments taking into account the approval status of anticancer drugs in Germany. Key features of the pipeline include ease of deployment and maintenance, as well as the use of open source software and modularity.

Implementation: Standardized workflows that provide automatic variant calling and annotation of genetic variants for clinical interpretation are beneficial in terms of performance and less error prone [1]. The open source tool GEMINI [2] already contains a number of relevant annotation databases and simplifies their installation. A workflow has been planned that uses GEMINI to support the annotation and identification of clinically relevant somatic variants. By importing into the cBioPortal open source visualization tool [3], the annotated variants should be displayed graphically to facilitate the assessment of the data. However, the GEMINI output has to be converted to the Mutation Annotation Format (MAF) prior import to cBioPortal.

A Python script accepts the GEMINI output and transforms the data into MAF. To ensure that the same program can be used, should the output of GEMINI or the import requirements for cBioPortal change, particular attention has been paid to the generic aspect. For the annotation pipeline in this case, this means that columns and rows can be transposed, the column names of the input can be automatically renamed, and the content mapped to cBioPortal. For example, in case of the Variant Classification two columns from GEMINI are mapped to one column in cBioPortal. In order to meet the requirements of a molecular tumor board platform, new features are designed such as displaying relevant anticancer drugs and the associated approval status in Germany.

Discussion: The functional interaction between genes and drugs as well as the approval status of anticancer drugs in Germany is relevant and should complement the visualization in cBioPortal. The knowledge base OncoKB [4], which is already integrated into cBioPortal, links actionable genes to anticancer drugs. The approval status of anticancer drugs in Germany must be supplemented for each drug in OncoKB. Continuous updates, such as approval status reviews, are challenging and should be automated as far as possible and stored in a database. These enhancements, in combination with the annotation pipeline and cBioPortal, are intended to provide assistance for implementing molecular tumor boards at the MIRACUM partner sites and beyond. To facilitate the delivery and installation of the annotation pipeline, relevant tools will be packaged in application containers. Together with the MIRACUM partners, the annotation pipeline will be further extended and evaluated. In the future, the possibility of exchanging annotation data between the individual partner sites should be defined.

MIRACUM is funded by the German Ministry for Education and Research (BMBF), FKZ 01ZZ1801C.

The authors declare that they have no competing interests.

The authors declare that an ethics committee vote is not required.