Implicit Correlation Intensity Mining Based on the Monte Carlo Method with Attenuation

Abstract

Rapid development of computer and network technology has greatly promoted the biological information science. People have made satisfactory achievements in the study of high-throughput interaction map and pathogenic gene identification, and have been able to verify the candidate associations between genes and disease. However, a large amount of implicit knowledge between diseases, symptoms and genes have not been discovered. With the arrival of the age of big data, the number and variety of biomedical data sets have had a huge breakthrough. The rapid growth of biomedical big data provides the possibility of discovering biomedical implied relationship and assessing the strength association between entities. This paper puts forward an implicit association mining algorithm combining Monte Carlo method with the Newton's law of cooling. The algorithm synthesizes path, known-correlation intensity and dynamic changes of associated network topology. It can effectively find out potential and meaningful association between biomedical entities, and can evaluate the strength of the association based on probability.