Identification Of Stress Responsive Crucial Genes In Brassica Napus Based On Hidden Markov Random Field Model

The oil content of Brassica napus(B.napus)can reach more than 50%.It is an important economic oil crop and widely planted all over the world.However,the harsh environment seriously affect the growth and yield of B.napus(such as low temperature,high temperature,drought,salt,ABA).Drought stress is a particularly unfavorable ob-stacle during the growth and development of B.napus.Therefore,it is of great significance to study the physiological response mechanism of B.napus under drought condition.It is a good tool to use the B.napus transcriptome data to construct a gene co-expression network to study the crucial genes of B.napus that responded to stress.According to the principle of guilt by rewiring,we uses the transcriptome data of B.napus under drought treatment in this thesis.This principle assumes that the genes more likely undergo rewiring under the stress condition.Then,based on the co-expression network,the hidden markov random field(HMRF)model is used to analyze the drought response genes in B.napus.First of all,the transcriptome sequencing data used in this thesis was obtained from our research team.For B.napus,with three biological replicates for each conditions,We consider five treatments,including low temperature,high temperature,drought,salt,and ABA.Totally 18 samples were sequenced using Illumina Hiseq 4000 platform,including 15 treated samples and 3 non-treated control samples.This thesis mainly uses the tran-scriptome data of 6 samples under drought stress condition and non-treated control condition.Totally 2590 genes that to under drought stress condition were screened from the transcriptome data.Secondly,based on the selected drought response genes,the t-test was used to obtain the similarity matrix between genes,and then we construct the co-expression network among genes based on this matrix.Subsequently,we use the Fisher's transformation to establish the rewiring network.Further use of HMRF,the co-expression network of the gene and the p data from the t-test of B.nopus were integrated.Finally,we identified 472 crucial genes of B.napus under drought stress.Finally,we use the PageRank,betweenness centrality,and the HMRF to calculate the score of the crucial response genes and then compare the results by the ROC curve.The results show that,the HMRF can better identify the response genes from the tran-scriptome data of B.napus under drought stress condition.Further,through the GO enrichment analysis of the response genes of B.napus,we found that the response genes identified by the HMRF were enriched in the biological processes associated with the drought stress response of the plants.After that,the network and the co-expression net-work will be reconnected.By comparative analysis the rewiring network and c--expression network,it was found that the response genes were more likely to rewire in the rewiring network;the rewiring network contained more information,and the HMRF made full use of various biological information.The associated investigation can be not only beneficial to people's understanding of the physiological process of B.napus under stress,but also beneficial to the improvement of B.napus variety.