networkGWAS

networkGWAS applies network-based genome-wide association analysis to aggregate effects of multiple genetic markers across genes, pathways, and subnetworks to detect genotype–phenotype associations in structured populations.

Key Features:

• Network-based aggregation: Aggregates effects of multiple genetic markers to test entire genes, pathways, and subnetworks for association with phenotypes.
• Mixed models and neighborhood aggregation: Employs mixed models combined with neighborhood aggregation to model interactions within gene networks.
• Population structure correction: Incorporates mechanisms to correct for population structure in structured populations.
• Multiple-testing calibration: Uses circular and degree-preserving network permutation schemes to obtain well-calibrated p-values and control false positives.
• Integration with gene-based GWAS: Systematically combines gene-based GWAS with biological network information.
• Computational efficiency: Implements a computationally efficient framework for network-based GWAS.

Scientific Applications:

• Detection of known associations: Demonstrated ability to detect known associations using semi-simulated common variants from Arabidopsis thaliana and simulated rare variants from Homo sapiens.
• Exploration of biological processes: Identifies neighborhoods of genes involved in stress-related biological processes in Saccharomyces cerevisiae.
• Complex trait analysis: Applicable to studies of complex traits influenced by interactions among multiple genes or pathways.

Methodology:

Integrates mixed models with neighborhood aggregation, combines gene-based GWAS with biological network information, applies circular and degree-preserving network permutation schemes for multiple-testing calibration, and includes mechanisms to correct for population structure within a computationally efficient framework.