Research On Signal Transduction Pathway Reconstruction And Component Identification

Biological signal transduction is to induce relevant gene expression, control the cell division, and determine the prognosis of the cells via a series of complex signal transmission process. Pathway is a basic module or local subnetwork which completes a certain biological function. Signal transduction pathway is a transmission channel which is formed by interactions among a series of components in the process of signal transmission, including receptor, protein kinase, G-protein and second messengers. It is a kind of molecular channel by which cell can receive and transduce external stimulation, and then regulate metabolic reactions and gene expression. Consequently, a certain endpoint biological function can be accomplished. Most biological functions of a cell are coordinated by the signal transduction pathways and regulatory mechanisms,and then cells respond to and accommodate oneself to an ever-changing environment.Research on signal transduction pathway is an essential means to understand the complex life process and the pathogenesis of disease. Only realizing pathway structure and function in biological organism clearly can the basic mechanism and process of life activity be understood in the level of cell molecule. On the basic of these cognition, the pathogenesis of disease and preventive and diagnostic approach can be known, and then effective medicine design can be made, finally, individual medical treatment can be implemented. Therefore, the research on signal transduction pathway becomes one of hotspots in the present bioinformatics.In this paper, a novel method of signal transduction pathway reconstruction is proposed which is based on HMM model. In the method, data basis of signal transduction pathway reconstruction is that all the components in the pathway are known, but relationships and directions among the nodes are partly missed. These data could belong to multiple data type, and they could be the results of biological experiments, text mining or high-throughout data analysis. HMM is taken as a signal transduction pathway construction model, while the algorithm of optimizing model parameters is constructed based on EM algorithm in the method. In the process of reconstruction, the data with missing values are used as samples and the experiments are completed by taking examples of PKA and MAPK pathway. Finally, based on the cellsignalling database, the method is testified, whereafter, the comparison analysis of accuracy between above result and the result in the literature is made. The actual effect of the reconstruction experiments showed that the accuracy of regulation relationship was improved, and further, regulation direction was also predicted. It is very important to understand the functions of signal transduction pathway and relational life phenomenon.Based on the research production, a new approach is proposed in which the priori knowledge is fused again. And it is proposed that S-DDI information is regarded as a kind of indirect priori knowledge. By the research on discovery method of S-DDI, a prediction algorithm of S-DDI is given in the paper in which ILP algorithm is utilized to discover authentic signaling domain-domain interactions from diverse genomic and proteomic databases combining with signal domain information. Consequently, the experiment is made by taking an example of MAPK pathway as well. And then, the accuracy analysis is completed while the comparison analysis in sensitivity and specification are finished. The actual effect showed that the prediction's accuracy was improved further by fusing priori knowledge. These productions have impulsed research on signal transduction pathway to develop in the direction of fine and detail.It is one of the key problems in the field of signal transduction pathway that the way of discovering the components in the pathway is developed on the basis of high-throughout data. In this paper, a kind of FCM clustering algorithm with two stages is proposed and the comparison analysis experiments of differentially expressed genes identification are completed in which a variety of SAM algorithms are utilized. These works found the basis for identifying the pathway components in signal transduction pathway.