ATAV

ATAV stores variant calls and per-site coverage data from whole-exome and whole-genome sequencing in a centralized database to address joint-calling and enable diagnostic (trio and singleton) analyses and rare-variant collapsing for disease-association discovery.

Key Features:

• Centralized variant and coverage database: Stores variant calls and per-site coverage data for all samples to address joint-calling and enable cohort-wide queries.
• Scalability: Demonstrated ability to analyze datasets comprising over 110,000 individuals.
• Diagnostic analysis support: Supports diagnostic analyses for trios and singletons to identify candidate causative variants.
• Rare-variant collapsing: Implements rare-variant collapsing analyses for discovery of disease-associated genes and variants.
• Modular architecture: Modularized framework that enables extensibility and continuous development of analysis components.
• Reproducibility logging: Records runtime logs that capture analysis provenance and enable full reproducibility.
• Aggregated sample summary data: Contains summary-level data for over 40,000 samples, including cases and controls from diverse ancestries, with phenotype categories, predicted ancestry, gender, and quality metrics.
• Efficient querying: Enables efficient queries of the comprehensive variant and coverage dataset to support downstream analyses.

Scientific Applications:

• Diagnostic variant identification: Identification of disease-causing variants in clinical and research settings using trio and singleton analyses.
• Disease-gene discovery: Discovery of disease-genes through rare-variant collapsing analyses, demonstrated on datasets with more than 20,000 samples.
• Rare-variant association studies: Detection of rare-variant associations in complex diseases using collapsing approaches.
• Ancestry-aware analyses: Use of aggregated summary-level data and predicted ancestry to support population- and ancestry-aware variant interpretation.

Methodology:

Stores variant calls and per-site coverage in a centralized database, enables efficient cohort-wide queries, performs diagnostic (trio/singleton) analyses and rare-variant collapsing, and records runtime logs for reproducibility.