Protein
Identifiers and Primary Sources
The IGVF Catalog has protein nodes collection with Ensembl Protein IDs (e.g.,ENSP00000384707) as identifiers from the following GENCODE releases:
- Human: GENCODE v43
- Mouse: GENCODE vM36
P49711), protein names, and Database cross-reference (dbxref). The data was imported from the following UniProt files (downloaded in May 2024):
- Human: uniprot_sprot_human.dat.gz, uniprot_trembl_human.dat.gz
- Mouse: uniprot_sprot_rodents.dat.gz, uniprot_trembl_rodents.dat.gz
Additional Protein Edge Collections
Complexes-Proteins Edges
Data protein complexes are imported from Complex Portal at the EBI.Motifs-Proteins Edges
Transcription Factor (TF) Binding Motifs represented by Position Weight Matrices (PWMs) defining the DNA binding preferences for transcription factors from HOCOMOCO v11.Variants-Proteins Edges
Protein-Protein Edges
Protein-protein interactions from BioGRID and IntAct includes direct physical interactions and functional genetic associations.Coding variants
Identifiers and Primary Sources
The IGVF Catalog has coding variants collection using identifiers combining Ensembl Transcript ID and HGVS nomenclature for both protein and cDNA changes (e.g.,HK1_ENST00000298649_p.Arg380Gly_c.1138C>G). Each coding variant is also mapped to its genomic location using standardized SPDI identifiers.
The primary source is dbNSFP v5.1, which aggregates scores from dozens of functional prediction algorithms (e.g., SIFT, CADD) for all possible non-synonymous single-nucleotide variants (nsSNVs) and splice-site variants in the human genome. Additionally, predictions on coding variants functions from Mutpred2 and ESM-1v for every possible single amino acid change in human genes are imported as part of the IGVF project. As part of this import process, all possible genetic indels that could lead to each specific amino acid change were enumerated and included.
Coding Variant Edges
Variants-Phenotypes-Coding variants Edges
- IGVF SGE assays: CRISPR-based assays that systematically introduce and evaluate the functional impact of hundreds to thousands of variants in PALB2, CTCF, RAD51D, SFPQ, XRCC2, and BRCA2.