Turbo Data Integration for Uncovering Gene Networks

Data integration is an important task in the gene network research. In this paper, the integration of two microarray datasets for uncovering gene networks is considered. The premise of data integration is that different datasets all carry information about the common networks of interest. Thus, through data fusion, the authors hope to combine information from different sources to gain improved understanding about the networks. Inspired by the similarity between the data integration problem and turbo decoding in wireless communications, a Bayesian data integration framework based on a turbo approach was proposed. The proposed algorithm was tested on two microarray datasets of 10 genes in the yeast cell cycle