Research On Evaluation Of The Node Importance In Gene Regulatory Network Via A Pseudo Knockout Index

Gene regulation plays an important role in cell life activities.It is a dynamic process.In the process of biological development,the interaction between genes is constantly changing,thus control the gene expression level.Gene regulatory network(GRN)refers to the set of gene-gene interactions,corresponding to regulatory relationships through their products,and the interactions in the GRN represent such rules.GRNs provide a possible framework to measure the importance of nodes.The evaluation of important nodes in GRN can effectively reveal the key genes in the network and discover their functional implications severing as key players in biological processes,such as master regulator and driver gene.At present,most of the existing methods are to evaluate a single node in a complex network,and they are mainly based on static network structure.In addition,it should be pointed out that the influence of gene node combinations in GRN is not simply additive.Genes play multiple functions such as cooperation,confrontation and mutual substitution through interaction.The discovery of important gene combinations,especially the combination of transcription factors,is of great significance to mining functional gene sets and revealing the synergistic regulation mechanism.In recent years,with the rapid development of sequencing technology,especially single-cell RNA sequencing(scRNA-seq)technology,high-throughput quantification of gene expression in single cells can be achieved.This provides a basis for identifying the key regulators and their collaboration in the process of biological development and disease pathogenesis.This paper proposed an approach called Pseudo Knockout Importance(PKI)to identify key genes and sets from gene regulatory networks.The specific GRN was constructed by using time course scRNA profiles combined with prior regulatory knowledge.Ordinary differential equation model was constructed to describe the dynamic regulation process.The importance of nodes and node sets in GRN was evaluated by pseudo gene knockout method,and important gene sets were discovered by mathematical programming method.In this paper,the method is applied to the core GRN of human embryonic stem cells(ECSs)to discover the key regulatory nodes in the network.The main contents are as follows:(1)Reconstruct ESC GRN during stem cell differentiation based on prior knowledge and time-course gene expression data.Then use ordinary differential equation(ODE)model to describe its dynamic regulation process.(2)Design gene pseudo knockout experiments and define PKI score evaluation criteria based on coefficient of determination,the consistency of gene expression data and ODE model.For key gene combinations,PKI is derived as a combinatorial optimization problem of quantifying the in silico knockout effects.Heuristic algorithm was used to solve the problem quickly.From the view of statistics and optimization,the important nodes and node-sets are given.(3)In order to verify the effectiveness and superiority of PKI method,PKI-based importance ranking of single node is compared with 12 other existing methods,and the minimum dominant set method is employed for comparison analysis.In addition,functional enrichment analysis is performed on genes with high order,and key transcription factors in ESCs development are found in the literature to verify the key gene combinations discovered by PKI method.These results demonstrate the reliability of the findings.