An Approach of Gene Regulatory Network Construction Using Mixed Entropy Optimizing Context-Related Likelihood Mutual Information

We propose a gene regulatory network construction method based on mixed entropy optimization context-relevant likelihood mutual information (MEOMI). Firstly, the mutual information between genes was calculated based on the combination of two entropy estimators, and then the distribution was optimized by the likelihood algorithm (CLR) to eliminate some indirect regulation relations, and the initial gene regulation network was obtained. In order to reflect the relationship between the complex interactions between genes and eliminate redundant edges in a gene regulatory network, this method is given by calculation of multiple gene under the conditions of the interaction between the two genetic information (CMI2) to adjust the initial gene regulatory network, continuously updated network to perfect the indirect regulating the relationship between genes, reduce the mutual information overestimated the impact of the strength of direct regulation.