A combinatorial searching method for detecting a set of interacting loci associated with complex traits

Area of Science:

• Genetics
• Biostatistics
• Complex disease research

Background:

• Complex diseases arise from interactions between multiple genes and environmental factors.
• Current genetic association studies often analyze one gene or marker at a time, potentially missing crucial multilocus effects.
• Identifying interacting genetic units is a significant challenge in understanding trait variability.

Purpose of the Study:

• To develop and evaluate a novel computational method, the Combinatorial Searching Method (CSM), for detecting sets of interacting genetic loci associated with complex traits.
• To address the limitations of single-locus analysis by considering multilocus interactions.
• To identify specific gene-gene interactions influencing systolic and diastolic blood pressure.

Main Methods:

• The Combinatorial Searching Method (CSM) was developed, employing a filtering step to identify candidate locus-sets.
• A new objective function utilizing cross-validation and multi-locus genotype partitioning was introduced to evaluate candidate sets.
• A permutation procedure was used to assess the statistical significance of identified locus-sets.

Main Results:

• Simulation studies demonstrated the CSM's capability to detect high-order gene-gene interactions.
• Application to real data revealed a four-locus interaction model as the best predictor for systolic blood pressure (SBP) (p=0.033).
• A two-locus interaction model was identified as the best predictor for diastolic blood pressure (DBP) (p=0.045).

Conclusions:

• The CSM is a powerful tool for identifying complex gene-gene interactions that contribute to disease risk and trait variability.
• The method successfully identified significant multilocus genetic associations for blood pressure traits.
• This approach offers a promising avenue for mapping genes involved in complex diseases.