VariantMatcher and the variant-level matching network to facilitate variant classification and disease gene discovery.
Laboratory Genetics and Genomics
-
Primary Categories:
- Genomic Medicine
-
Secondary Categories:
- Genomic Medicine
Introduction:
VariantMatcher (variantmatcher.org) is a freely available web-based tool developed to connect individuals worldwide interested in a specific variant. It enables sharing de-identified variant-level and phenotypic data from research participants to discover disease-causing variants and genes. The database contains rare (MAF < 1% in gnomAD) (Karczewski et al., 2020), nonsynonymous SNVs identified in 6,827 VCF files (957,249 unique variants) of affected and unaffected individuals sequenced as part of multiple projects along with their phenotypic information. Users can query the database upon registration and administrator approval. VariantMatcher is a founder member of the Federated Variant-Level Matching GA4GH Implementation Forum project and is now connected to the Franklin by genoox and MyGene2/Gene2MP databases through the Beacon v2 API. VariantMatcher is also connected to the Beacon Network.
As of 1 November 2024, VariantMatcher had 1,254 users from 63 countries. 21,261 variants were queried and 562 variants matched to 5,003 individuals in VariantMatcher. Additionally, 65,821 variants were queried through the Beacon Network, and 1,586 variants were matched to 27,083 individuals in VariantMatcher.
Methods:
To date, most matches have supported the more precise variant classification of variants initially classified as VUSs, enabling better, more precise patient care. To investigate further the use of VariantMatcher and its benefit in variant classification, we analyzed VUSs identified by Diagnósticos da América S.A., DASA, a laboratory in Brazil, between 2018 and 2021 and present in VariantMacher.
Results:
1,679 classified as VUSs in the DASA’S database were further investigated for matches using the Beacon Network to query VariantMatcher, PhenomeCentral, and Gene2MP. 1,137 variants were found in gnomAD, while 542 were not. Of these 542, 468 were present in a Beacon Network-connected database, including 80 that were present in VariantMatcher. Next, we investigated these 80 variants in the VariantMatcher database. A total of 63 variants were found to have a phenotypic overlap, and the phenotypes were further investigated. Based on the phenotype analyses of the individuals with the variants of interest, the classification of these variants was reviewed by two medical geneticists. The review of the classification took into consideration the zygosity of the variant, the mode of inheritance of the disease caused by the gene, the age of onset of the disease, the clinical features of the affected individuals, if the variant was present in unaffected individuals in VariantMatcher, and the segregation of the variants in other family members available in VariantMatcher. Of these 63 VUSs, 23 were in genes that cause autosomal or X-linked dominant diseases. By analysis of the variant matches’ phenotypes, four variants were reinterpreted as 'benign' and 14 as 'likely benign’; five remained classified as VUSs. Fifteen were in genes that cause late-onset diseases, mainly cancers. One variant was reinterpreted as 'benign' and one as 'likely pathogenic’; 13 remained classified as VUSs. Twenty-five were in genes that cause autosomal or X-linked recessive diseases and all of them remained classified as VUSs.
Conclusion:
To date, most of the VariantMatcher matches helped to rule out the queried variant as causative for the phenotype being investigated, mainly for the autosomal or X-linked dominant diseases with early onset and complete penetrance. However, before ruling out the variant as a candidate, users should consider incomplete penetrance, variable expressivity of the phenotype, and age of onset of the phenotypes under investigation (Wohler et al., 2021).
VariantMatcher has been widely used around the globe and has enabled researchers, health care providers, and patients to share their phenotypic and genomic data and build new connections to increase the discovery rate of novel etiologies for diseases of unknown causes providing insights into both human biology and disease.
VariantMatcher (variantmatcher.org) is a freely available web-based tool developed to connect individuals worldwide interested in a specific variant. It enables sharing de-identified variant-level and phenotypic data from research participants to discover disease-causing variants and genes. The database contains rare (MAF < 1% in gnomAD) (Karczewski et al., 2020), nonsynonymous SNVs identified in 6,827 VCF files (957,249 unique variants) of affected and unaffected individuals sequenced as part of multiple projects along with their phenotypic information. Users can query the database upon registration and administrator approval. VariantMatcher is a founder member of the Federated Variant-Level Matching GA4GH Implementation Forum project and is now connected to the Franklin by genoox and MyGene2/Gene2MP databases through the Beacon v2 API. VariantMatcher is also connected to the Beacon Network.
As of 1 November 2024, VariantMatcher had 1,254 users from 63 countries. 21,261 variants were queried and 562 variants matched to 5,003 individuals in VariantMatcher. Additionally, 65,821 variants were queried through the Beacon Network, and 1,586 variants were matched to 27,083 individuals in VariantMatcher.
Methods:
To date, most matches have supported the more precise variant classification of variants initially classified as VUSs, enabling better, more precise patient care. To investigate further the use of VariantMatcher and its benefit in variant classification, we analyzed VUSs identified by Diagnósticos da América S.A., DASA, a laboratory in Brazil, between 2018 and 2021 and present in VariantMacher.
Results:
1,679 classified as VUSs in the DASA’S database were further investigated for matches using the Beacon Network to query VariantMatcher, PhenomeCentral, and Gene2MP. 1,137 variants were found in gnomAD, while 542 were not. Of these 542, 468 were present in a Beacon Network-connected database, including 80 that were present in VariantMatcher. Next, we investigated these 80 variants in the VariantMatcher database. A total of 63 variants were found to have a phenotypic overlap, and the phenotypes were further investigated. Based on the phenotype analyses of the individuals with the variants of interest, the classification of these variants was reviewed by two medical geneticists. The review of the classification took into consideration the zygosity of the variant, the mode of inheritance of the disease caused by the gene, the age of onset of the disease, the clinical features of the affected individuals, if the variant was present in unaffected individuals in VariantMatcher, and the segregation of the variants in other family members available in VariantMatcher. Of these 63 VUSs, 23 were in genes that cause autosomal or X-linked dominant diseases. By analysis of the variant matches’ phenotypes, four variants were reinterpreted as 'benign' and 14 as 'likely benign’; five remained classified as VUSs. Fifteen were in genes that cause late-onset diseases, mainly cancers. One variant was reinterpreted as 'benign' and one as 'likely pathogenic’; 13 remained classified as VUSs. Twenty-five were in genes that cause autosomal or X-linked recessive diseases and all of them remained classified as VUSs.
Conclusion:
To date, most of the VariantMatcher matches helped to rule out the queried variant as causative for the phenotype being investigated, mainly for the autosomal or X-linked dominant diseases with early onset and complete penetrance. However, before ruling out the variant as a candidate, users should consider incomplete penetrance, variable expressivity of the phenotype, and age of onset of the phenotypes under investigation (Wohler et al., 2021).
VariantMatcher has been widely used around the globe and has enabled researchers, health care providers, and patients to share their phenotypic and genomic data and build new connections to increase the discovery rate of novel etiologies for diseases of unknown causes providing insights into both human biology and disease.