The development of statistical pathway analysis methods has focused on testing individual main effects of genes in a pathway on disease. However, gene-gene interactions can also play an important role in complex disease etiology. We developed a pathway analysis method based on a protein-protein interaction network to account for gene-gene interactions in a pathway. We used simulations to evaluate the type I error and power for the method. Our simulation results suggest that the method has correct type I error rates, and can be powerful in the identification of the effects of gene-gene interaction in pathways under different scenarios. The method has been implemented into an efficient software package with C++.
Date:
2013-04
Relation:
2013 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB). 2013 Apr:238-241.
