Revealing Disease Related Interactions by Correlation Analysis

摘要

The computational identification of disease related lesions is still a key open problem in biomedicine and systems biology. Dysregulated interactions may be an important reason that causes disease. In this paper, we aim to identify dysregulated interactions so as to elucidate the mechanism of disease in a systematic manner. Specially, we present a method to detect which protein-protein interactions or genetic interactions are downregulated or upregulated due to disease process. The proposed method was applied to a human molecular interaction network and a prostate cancer microarray dataset to reveal dysregulated interactions. The enrichment analysis of cancerous genes and disease related GO terms in identified dysregulated interactions shows that the identified dysregulated interactions are disease related, which verifies the effectiveness of our method.