An automatic model building for screening functional foods with GP

Health and disease are one of the most important issues for us. Especially, cancer is the most serious problem in recent decades. Some researchers focus on functional foods that have good effect for human body to prevent cancer. However, there are various kinds of foods in the world, and to develop a high-throughput screening method is anticipated. In this paper, we propose a method to estimate physiological activity of foods from protein expression levels by genetic programming when the constituent of foods is given to cells. Case studies show that our method improved accuracy to estimate anti-angiogenic activity than that by multidimensional regression analysis.