HOGImine

HOGImine discovers higher-order genetic interactions and integrates multiple variant encodings and biological priors to identify genetic meta-markers associated with complex phenotypes.

Key Features:

• Higher-Order Genetic Interactions: Considers higher-order interactions among genes to expand the class of discoverable genetic meta-markers beyond lower-order methods.
• Multiple Encodings for Genetic Variants: Supports multiple genotype encodings (e.g., recessive and dominant) rather than requiring binary encodings.
• Integration of Biological Priors: Incorporates priors such as protein-protein interaction networks, genetic pathways, and protein complexes to restrict the search space.
• Enhanced Statistical Power: Demonstrates substantially higher statistical power in experimental evaluations to detect phenotype-associated genetic mutations missed by lower-power methods.
• Efficient Computational Strategy: Employs a more efficient search strategy to reduce runtime when searching higher-order gene interactions, enabling analyses at larger scales.

Scientific Applications:

• Complex disease genetics: Identify combinations of genetic variants and higher-order interactions associated with multifactorial diseases.
• Developmental biology: Detect genetic interaction patterns underlying morphological traits.
• Evolutionary genetics: Investigate combinatorial genetic architectures relevant to evolutionary divergence and trait evolution.
• Hypothesis-driven network analysis: Test and prioritize hypotheses within known biological networks such as protein complexes and pathways.

Methodology:

Considers higher-order gene interactions, supports multiple genotype encodings, integrates biological priors (protein-protein interaction networks, genetic pathways, protein complexes) to restrict the search space, and employs an efficient search strategy to reduce runtime.