BALL-SNPgp

BALL-SNPgp maps non-synonymous single nucleotide polymorphisms (nsSNPs) from VCF or tabular Next-Generation Sequencing (NGS) data onto Protein Data Bank (PDB) structures and uses the Biochemical Algorithms Library (BALL) to integrate structural, database, and in silico prediction information for pathogenicity assessment.

Key Features:

• Input formats and structural mapping: Accepts variant call format (VCF) files or tabular NGS input and retrieves corresponding Protein Data Bank (PDB) structures to map variants.
• Visualization of mutated residues: Highlights mutated residues within three-dimensional protein structures to show precise spatial locations of nsSNPs.
• Hierarchical bottom-up clustering: Performs hierarchical bottom-up clustering of nsSNPs based on spatial proximity within the 3D protein structure.
• Database and prediction integration: Integrates pathogenicity and annotation data from ClinVar and HUMSAVAR and incorporates I-Mutant2.0 for in silico protein stability/pathogenicity predictions.
• Pathogenicity aggregation: Aggregates pathogenicity information from multiple sources and presents it alongside structural mappings for comparative assessment.
• Functional site prediction: Predicts binding pockets and other functional sites within protein structures to evaluate potential effects of nsSNPs on interactions and function.

Scientific Applications:

• Clinical variant interpretation: Supports interpretation of nsSNP pathogenicity in clinical genomics and high-throughput NGS variant datasets.
• Detection of spatial variant clusters: Identifies spatially proximate clusters of nsSNPs that may indicate collective effects on protein function or pathogenicity.
• Cardiomyopathy research: Has been applied to cardiomyopathy studies to detect significant accumulations of variants in genes such as JUP, VCL, and SMYD2.

Methodology:

Imports variant data from VCF or tabular input, retrieves corresponding PDB structures, visualizes 3D structures with highlighted nsSNPs, applies hierarchical bottom-up clustering based on spatial proximity, integrates ClinVar and HUMSAVAR annotations and I-Mutant2.0 predictions, and predicts binding pockets and other functional sites.