ADDO

ADDO detects and characterizes additive and non-additive quantitative trait loci (QTLs) from whole-genome sequence data to quantify additive, dominance, partial dominance, and overdominance effects while accounting for population structure and relatedness.

Key Features:

• Mixed-Model Transformation: Employs mixed-model transformation that accounts for additive and dominant genetic covariance to control for population structure and unequal relatedness among individuals.
• Decomposition of SNP Effects: Decomposes single nucleotide polymorphism (SNP) effects into additive, partial dominance, dominance, and overdominance categories.
• Efficient Computation: Uses matrix multiplication approaches to accelerate genome scans (e.g., a scan of ~20 million markers across 836 individuals completed in ~8.5 hours on 10 CPUs as reported).
• Comprehensive Analysis Pipeline: Provides a systematic pipeline for characterizing QTLs in whole-genome sequence data and complements additive-focused GWAS by explicitly modeling non-additive effects.

Scientific Applications:

• Complex trait dissection: Enables detection and classification of additive and non-additive genetic contributions to complex trait variation.
• Detection of dominance-driven QTLs: Identifies QTLs driven by dominance and other non-additive effects that may be missed by additive-only GWAS models.
• Whole-genome sequence QTL characterization: Characterizes QTLs using large-scale genomic datasets with millions of markers and hundreds to thousands of individuals.
• Validation in simulated and real datasets: Has been applied to simulated data and real-world examples, including outbred rat studies where significant dominant QTLs were detected that were not found by additive models.

Methodology:

Handles large-scale genomic data (millions of markers from hundreds to thousands of individuals); applies mixed-model transformations that account for additive and dominant genetic covariance to control population structure and relatedness; decomposes SNP effects into additive, partial dominance, dominance, and overdominance; leverages matrix multiplication techniques to expedite genome-wide scans.