Robust Gene Network Topology Construction Based on the Evolutionary Algorithm and Artificial Neural Network

Abstract

Design and implementation of automatic gene regulatory network are essential to construct and analyze the complex biological system. The recent study shows that Darwinian evolution can gradually develop higher topological robustness. In these consequences, this thesis presents an integrated scheme to simulate gene expressions dataset for identifying network topologies to find the robustness based on an evolutionary approach and artificial neural network. The final outcome is the most robust topology from a gene regulation dataset. The proposed method was verified using randomly sampled parameter spaces and threshold are generated by the network itself. Here, final result shed lights on the relationship among genes and corresponding transcription factors. Transcription factors are combined to specify the on-and-off states of genes. This binding form a regulatory network and constituting the wire diagram for a cell. The proposed network shows the whole combinatorial and co-association of transcription factors, co-relation and the robustness of human genes. Therefore, this research will play a crucial role in interpreting personal genome sequences and understanding basic principles of human health evolution in near future.