SPDI

SPDI represents and normalizes genetic sequence variants using NCBI's four-attribute Sequence Position Deletion Insertion model to enable consistent normalization, projection, and aggregation of variants for genomic analyses.

Key Features:

• Four-Attribute Representation: Variants are defined by sequence, position, deletion, and insertion attributes applicable to both nucleotide and protein variants.
• Normalization and Format Conversion: Web services convert between HGVS, VCF, and SPDI formats and the NCBI Variant Overprecision Correction Algorithm generates a unique Contextual Allele for normalized representation.
• Handling Repeat Regions: The model precisely defines the reference subsequence affected by a variant, including in homopolymer and other repeat regions.
• Projection Across Congruent Sequences: Variants are projected across congruent sequences using an alignment dataset that includes non-assembly RefSeq sequences (NM, NR, NG) and inter- and intra-assembly genomic sequences (NC, NT, NW).
• Canonical Allele Designation: Among projected Contextual Alleles, one Canonical Allele is designated—typically based on the latest assembly version—to represent the set of congruent sequences for aggregation.

Scientific Applications:

• Variation databases: Representation and aggregation of asserted sequence variants in NCBI resources such as dbSNP and ClinVar.
• Genomics research: Consistent variant normalization and projection to support genomic analyses, variant curation, and the study of genetic diseases and biological functions.

Methodology:

SPDI converts variant representations between HGVS/VCF/SPDI, applies the NCBI Variant Overprecision Correction Algorithm to produce Contextual Alleles, projects variants across congruent sequences using an alignment dataset (NM/NR/NG and NC/NT/NW), and defines precise operations to address discrepancies among variant callers and databases.