11. The complex nature of variant interactions in cancer requires updates to variant interpretation resources

Kilannin Krysiak

Kilannin Krysiak is an Assistant Professor at Washington University School of Medicine and Assistant Medical Director of the Cytogenomics and Molecular Pathology Laboratory. She received her PhD and completed her Laboratory Genetics and Genomics Fellowship at Washington University in St. Louis. During her postdoctoral work with Drs. Malachi Griffith and Obi Griffith, she was part of the original team that created the CIViC (Clinical Interpretation of Variants in Cancer; www.civicdb.org) knowledgebase and continues to be a lead curator and feature development consultant of that resource. As member of the ClinGen Somatic Somatic Working Group, she continues her international involvement in guideline development in the field of cancer molecular genetics.

Abstract

Kilannin Krysiaka, Susanna Kiwalaa, Adam C. Coffmana, Joshua F. McMichaela, Arpad M. Danosa, Jason Salibaa, Cameron J. Grisdaleb, Jake Leverc, Lana Shetaa, Shruti Raod, Alex H. Wagnere, Malachi Griffitha, Obi L. Griffitha

aWashington University School of Medicine, St. Louis, MO, USA; bCanada‚Äôs Michael Smith Genome Sciences Centre, Vancouver, BC, CA; cUniversity of Glasgow, Glasgow, GLG, UK; dGeorgetown University Medical Center, Washington, DC, USA; eNationwide Children’s Hospital, Ohio State University, Columbus, OH, USA

CIViC, the Clinical Interpretation of Variants in Cancer knowledgebase (www.civicdb.org), has provided structured data and an open-access curation interface to enable scientists and other community stakeholders to curate variant information from published literature since 2014. Since then, CIViC has continued to evolve to incorporate new community guidelines and user feedback. CIViC has been embraced by the community with >300 contributors submitting content that represents >3100 publications, >450 genes, and >3000 variants, and become the official knowledgebase supporting somatic cancer variant curation efforts for ClinGen. Here we present CIViC 2.0, the continued evolution of this resource with a significant redesign of both the data model and the user interface (UI). Enhancements to the UI improve the searchability, navigation and overall use of the web interface while changes to the underlying database structure allow for more complex variant types and relationships. One major component of this redesign is the support for Molecular Profiles, allowing complex variant relationships to be modeled in a structured format. Analysis of larger cohorts has advanced our knowledge of the prognostic and therapeutic impact of variant combinations in specific cancer types. For example, the co-occurence of both NPM1 W288fs and FLT3-ITD in acute myeloid leukemia indicates a better prognosis than FLT3-ITD alone. Although clinical guidelines have recognized such variant combinations for several years, few knowledgebases represent this level of complexity. In summary, CIViC 2.0 extends existing features that have engaged users in order to support the needs of the community with more complex and clinically-relevant capabilities.