2-B-P-092 〇亀淵 由乃¹、難波 里子¹、関谷 拓海¹、大谷 則子^1,2、岩田 通夫¹、山西 芳裕^1,2 Kamenofuchi Yoshino¹, Satoko Namba¹, Takumi Sekiya¹, Noriko Otani^1,2, Michio Iwata¹, Yoshihiro Ymanishi^{1,2 1}九工大・院情工・生化情、²名大・大学院情報学研究科 ¹Dep. Bio. Bioi., Facu. Comp. Sci. Syst. Eng., KyuTech, ²Grad. Info., Nagoya. Univ In recent years, drug combination therapy, which utilizes the synergistic effects of combining multiple drugs, has been attracting attention for medical treatment of multifactorial diseases such as cancer. The advantage of drug combination therapy is that it is expected to enhance therapeutic efficacy, but the disadvantage is that blind combination of drugs may cause harmful side effects. Therefore, it is necessary to identify the optimal combination of drugs. In this study, we develop a computational method to predict synergistic drug combinations from the viewpoint of regulation of therapeutic target molecules. We evaluate the coverage of a group of target molecules of the combined drugs, because the regulation of many diseases therapeutic target molecules may enhance therapeutic efficacy. We develop an algorithm to search for drug pairs with high coverage of therapeutic target proteins of each disease considering the potential target proteins of the drugs using machine learning models on various biomedical big data. The proposed method was applied to predicting drug combinations with synergistic effects for acute myeloid leukemia, chronic myeloid leukemia, colorectal cancer, and breast cancer. The predicted drug combinations were validated using clinical data. The proposed method is expected to contribute to the identification of optimal drug combinations for various diseases.