Modeling Protein-Signaling Networks with Granger Causality Test

摘要: The development of computational techniques to identify the gene networks, such as regulatory networks and protein–protein interaction underlying observed expression patterns, protein image data is a major challenge in analysis high-throughput data. Gene can be critical treatment complex diseases. Significant progresses have been made last few years characterizing interactions at genomic level [2, 6, 9], including methods for identifying interactions, modules occurring with high frequency genome, identification transcription motifs [3, 17, 19, 24]. Methods network reconstruction proposed based upon statistical Bayesian [16, 20, 21, 26], Boolean models [18], graphical Gaussian [8, 23], etc. key reconstructing causal relations among simultaneously acquired signals. Karen Sachs et al. [22] reconstruct signaling structure inference algorithm. However, several limitations. First, cost inferences usually very high, obtained results are not always accurate, comparison other reverse engineering approaches (see, example, [4]). Second, only applied pathways that they acyclic, whereas known rich feedback loops. One approach analyze causality between two signals examine if prediction one signal could improved by incorporating information