Sources of data in ClinVar

ClinVar is built from data provided by submitters. ClinVar is not staffed to review the literature to build content via curation. We gratefully acknowledge all those who contribute their data. This document cannot acknowledge all our contributors explicitly; our submitters page provides more explict attribution. This document summarizes some of the major categories of submitters, and how submissions in those categories are processed.

Practice guidelines/Professional societies

Groups that provide practice guidelines for specific variants may apply for recognition as such by ClinGen. If the application is approved, interpretations provided by that submitter will have precedence over all others, and will be represented with a review status of 4 stars. All submissions with that status can be retrieved by applying the filter Practice guideline.

Expert panels

Groups that review evidence about the clinical significance of variants may apply for recognition from ClinGen as an expert panel. If the application is approved, interpretations provided by that submitter will have precedence over those not from an expert panel or professional guideline, and will be represented with a review status of 3 stars. All submissions with that status can be retrieved by applying the filter Expert panel.

Documentation supporting each expert panel is provided on our FTP site.

Community projects

SCRP: The Sharing Clinical Reports Project – BRCA1 and BRCA2

SCRP stands for the Sharing Clinical Reports Project. It is a volunteer, grass-roots effort to encourage open sharing of genetic variant information. SCRP specifically aims to collect information on BRCA1/2 variants and make this information publicly available in ClinVar.

The problem of interpreting variants discovered through clinical testing in the BRCA1 and BRCA2 genes, as well as in many other genes involved in hereditary disorders, is one of the most critical problems facing clinical geneticists and their patients. In response to this need, the NIH and, in particular, NCBI has launched ClinVar to host fully anonymized genotype and phenotype information for all hereditary disease gene variants.

The goal of SCRP is to put variants in BRCA1 and BRCA2 that have been reported by Myriad Genetics, Inc. to cancer genetics or hereditary cancer clinics in the US since January 2006 into ClinVar. This date was chosen because it is when the company stopped sharing their variant information with the Breast Cancer Information Core (or BIC) database. The BIC was established as an open-access international collaboration for hosting an online BRCA1/2 mutation database, currently located at the National Human Genome Research Institute. All data obtained through the effort described here and submitted to ClinVar would automatically also be submitted to the BIC.

The future of molecular genetic testing depends on having data on the clinical validity of variants, e.g. their penetrance, if we are going to make any progress. This is especially true once large-scale whole exome or genome sequencing becomes more widespread and variants in thousands of disease genes are uncovered as incidental findings. There is a clear international consensus that we need openly accessible, variant databases for all disease genes, BRCA1 and BRCA2 included.
Please visit https://www.clinicalgenome.org/data-sharing/sharing-clinical-reports-project-scrp/ for more information on this effort and instructions for submitting variants.

To review the submissions via ClinVar, you can submit a query based on submitter name. Try it.

Genetic testing laboratories

Genetic testing laboratories have submitted data about alleles identified during their testing, with related literature that contributed to the interpretation.  They also have the option to provide information about how they assess the pathogenicity of any allele, the so-called Assertion Criteria. If a laboratory provides these assertion criteria, their submission will be represented with a review status of 1 star.

Locus-specific databases

Locus-specific databases have submitted information about the variants they have processed, often with links to the literature and summary data. ClinVar provides links back to those databases by allele. We look forward to expanding this data set.

Research studies

ClinSeq®, PubMed 22703879 was the first group to request ClinVar accessions pre-publication.

Semi-automatic data flows

OMIM

NCBI updates data from OMIM daily. This includes information about phenotypes, genes, and allelic variants. The content is used to maintain the query interface which directs results to omim.org, to integrate information about disorders and their related genes for Gene, GTR, and MedGen, and to identify allelic variant data for ClinVar. The description of the allelic variant is mined to extract complete HGVS expressions when text includes one nucleotide accession with version and  a representation of the sequence change, within parentheses.  As resources allow, the remainder of allelic variants are evaluated by curators to establish a molecular description.

GeneReviews

The molecular genetics section of novel and updated GeneReviews are evaluated for submission to ClinVar as part of GeneReviews' editorial processing. This includes any documents submitted as part of a GeneReview but not incorporated into the article explicity. ClinVar encourages authors of GeneReviews to submit information about variants to ClinVar directly. See how to submit.

Other public databases

UniProtKB

UniProt has submitted information about variants to NCBI databases, partially in conjunction with the RefSeqGene/LRG collaboration. The submissions are not comprehensive, but the links to the variant pages, e.g. http://web.expasy.org/variant_pages/VAR_069097.html, can be useful in bridging between nucleotide-specific and protein-specific information.

UniProt has also begun to connect AlleleIDs in ClinVar to their VAR identifiers.  By so doing, ClinVar is now reporting the HGVS expressions anchored on UniProt sequences that UniProt provided.  Links are also provided to the variant section of a UniProt record, e.g. P78363#VAR_008401.

Last updated: 2016-11-02T06:42:10-04:00