dbNSFP

dbNSFP provides comprehensive functional predictions and annotations for human nonsynonymous single-nucleotide variants (nsSNVs) and splice-site variants (ssSNVs) to support filtering and prioritization of variants from exome-sequencing and other genomic studies.

Key Features:

• Comprehensive Variant Database: dbNSFP v3.0 catalogs 82,832,027 nsSNVs and ssSNVs based on the human reference sequence annotated with GENCODE version 22 and includes dbscSNV with 15,030,459 potentially functional SNVs in splicing consensus regions.
• Functional Predictions and Annotations: Aggregates numerous prediction scores including MetaSVM, MetaLR, CADD, VEST3, PROVEAN, 4× fitCons, fathmm-MKL, DANN and the original v2.0 scores SIFT, PolyPhen-2 (2×), LRT, MutationTaster, MutationAssessor, and FATHMM.
• Allele Frequency Data: Integrates allele frequency information from cohorts including UK10K and the Exome Aggregation Consortium (ExAC).
• Updated Annotations: Version 3.0 updates SNV and gene functional annotations to improve accuracy and relevance of variant annotations.

Scientific Applications:

• Genetic Research: Facilitates identification and prioritization of disease-associated variants from exome and genome sequencing datasets.
• Personalized Medicine: Assists interpretation of variant impacts for individual genomes using aggregated prediction scores and annotations.
• Population Genetics: Enables analysis of allele frequency distributions across cohorts using integrated UK10K and ExAC data.

Methodology:

Compiled all potential nsSNVs and ssSNVs from the human reference genome, curated variant records, and integrated multiple prediction algorithms to annotate and score variant impacts.