A powerful score-based test statistic for detecting gene-gene co-association.

Area of Science:

• Genetics
• Biostatistics
• Computational Biology

Background:

• Genome-wide association studies (GWAS) explain limited heritability for complex diseases.
• The
• missing heritability
• may be due to gene-gene joint effects, including main effects and co-association.
• Gene-gene co-association considers the correlation between genes, not just traditional interactions.

Purpose of the Study:

• To propose a novel score-based statistic (SBS) for detecting gene-gene co-association.
• To address the gap in explaining complex disease heritability.
• To develop a gene-based method for analyzing correlated genetic loci.

Main Methods:

• Developed a novel score-based statistic (SBS) for gene-based analysis.
• Compared SBS performance against single nucleotide polymorphism (SNP)-based, principle component analysis (PCA)-based logistic regression, canonical correlations (CCU), kernel canonical correlation analysis (KCCU), partial least squares path modeling (PLSPM), and delta-square (δ (2)) statistics.
• Utilized simulation studies with varying sample sizes, causal SNP effects, and co-association levels.
• Applied the method to real-world rheumatoid arthritis (RA) data.

Main Results:

• The proposed SBS demonstrated superior performance compared to existing methods across various simulation scenarios.
• SBS showed better detection of gene-gene co-association under different conditions.
• Real data analysis of rheumatoid arthritis confirmed the practical advantages of SBS.

Conclusions:

• The score-based statistic (SBS) is a powerful and efficient gene-based method.
• SBS effectively detects gene-gene co-association.
• The method contributes to understanding the genetic architecture of complex diseases.