In:
Cancer Research, American Association for Cancer Research (AACR), Vol. 78, No. 13_Supplement ( 2018-07-01), p. 1290-1290
Abstract:
The Clinical Interpretations of Variants in Cancer (CIViC) database was designed as a fully open-access resource specifically focusing on the subset of cancer variants with well-defined clinical information associated to them and targeting a broad user base in cancer, including clinicians, researchers and patient advocates. CIViC follows a crowd-sourced curation model with expert moderation, and emphasizes transparency in that all changes, approvals, and discussion of content are tracked and displayed in the interface. Evidence items (EIDs) make up the fundamental unit of CIViC. EIDs connect a predictive, prognostic, diagnostic or predisposing clinical significance to a variant in the context of a specific disease, and each EID links back to specific published and peer-reviewed evidence. Multiple EIDs of high value can be used to make clinical assertions. While the CIViC variant is intentionally broad and can capture specific SNVs and Indels, while also admitting umbrella variants such as “EGFR mutation,” it is also anchored to a single gene entity (pulled directly from the NCBI Entrez database). Two drawbacks result from this strict association. First, clinical information associated with multiple genes is difficult to implement in knowledgebases like CIViC. In cases where two variants co-occur on the same gene, such as the erlotinib-sensitizing EGFR L858R and resistance mutation T790M, CIViC allows a “L858R and T790M” variant to be created under the EGFR gene. In other cases, such as FLT3 internal tandem duplication and DNMT3A mutation in AML, there is no CIViC variant that can capture this combination. In addition, non-gene entities important in cancer such as viral oncogene, microsatellite instability, or loss of chromosome region are not possible to represent. To address this, we are preparing CIViC V2, which will be implemented via modification of the existing database and web interface. The architecture around the EID as fundamental unit of CIViC will remain unchanged, but the CIViC variant will be re-envisioned, introducing the concepts of region and genotype. A variant will link to one or more regions, and region will be drawn from a list of types—genome, chromosome, rearrangement, gene and pathogen—which can be expanded when needed. The notion of complex genotype will be realized as a combination of variants drawn from multiple regions in either an additive manner (HER2 overexpression AND PIK3CA mutation) or employing more complex variant combinations using OR and NOT. Multicomponent genotypes will point back to EID collections associated with each individual component, while also enabling users to write EIDs unique to that genotype, which can, in turn, be used to create genotype clinical assertions. This new scheme will be presented via refactoring of the UI, and allow for a highly flexible concept of region and genotype reflecting the growing understanding of the combinatorial nature of variants in cancer. Citation Format: Arpad Danos, Kilannin Krysiak, Alex Wagner, Susanna Kiwala, Joshua McMichael, Adam Coffman, Erica Barnell, Yang-Yang Feng, Benjamin Ainscough, Cody Ramirez, Malachi Griffith, Obi Griffith. Expanding the CIViC variant to complex combinations of regions in the cancer genome [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 1290.
Type of Medium:
Online Resource
ISSN:
0008-5472
,
1538-7445
DOI:
10.1158/1538-7445.AM2018-1290
Language:
English
Publisher:
American Association for Cancer Research (AACR)
Publication Date:
2018
detail.hit.zdb_id:
2036785-5
detail.hit.zdb_id:
1432-1
detail.hit.zdb_id:
410466-3
