Algorithm Research On Pattern Recognition Of Liver Cancer Gene Regulatory Network Based On Tree Structur

Hepatocellular carcinoma is a cancer with high mortality and high insidiousness.Most patients with hepatocellular carcinoma are diagnosed in the middle to late stages and have poor prognosis.Studying hepatocellular carcinoma gene regulatory networks is an important way to explore the generation,diagnosis and treatment of hepatocellular carcinoma.Network motifs are subgraphs that appear more frequently than in random networks,reflecting the generation pattern and corresponding functions of the original network.By analyzing the studies on gene regulatory networks of liver cancer in recent years,we found that the domestic and international studies on gene regulatory networks of liver cancer only target individual micro RNAs or transcription factors,and the studies on liver cancer from the perspective of complete micro RNA-transcription factor gene regulatory networks are not deep enough.The existing network motif detection tools have low detection efficiency and slow speed when applied to large sparse networks,which seriously affects the in-depth analysis of gene regulatory networks.In this paper,the sequencing data collected from hepatocellular carcinoma and normal liver tissues were analyzed using the Mi CPAR algorithm to predict the target genes of micro RNAs in them,and merged with the transcription factor-target gene pairs from the publicly available database GRNdb to obtain the complete micro RNA-transcription factor regulatory networks of hepatocellular carcinoma and normal liver tissues,containing 23409 and 22349 regulatory factors and 286637 and259190 regulatory relationships respectively.Meanwhile,this paper proposes a tree-based subgraph search algorithm and a subgraph isomorphism testing algorithm,which can rapidly detect multi-node network motifs of sparse networks.By comparing the experiments with existing network motif detection tools,the network motifs of eight networks of different sizes from different domains are identified,and the experimental results verify the reliability,rapidity and stability of this algorithm.In this study,the 3-and 4-node network motifs of micro RNA-transcription factor regulatory networks in hepatocellular carcinoma and normal liver tissue were examined using a tree-based subgraph search algorithm,and the results showed that the network patterns in hepatocellular carcinoma did not change,but the Z scores of feedforward loops became significantly higher.The network of mutual regulation of regulatory factors was more significant in hepatocellular carcinoma compared to the network in normal liver tissue and the influence of multiple input modules was reduced.This study speculates that it may be due to the occurrence of gene mutations in hepatocellular carcinoma,which leads to the disruption of regulatory relationships and affects the regulation between regulatory factors,which eventually manifests as abnormal gene expression and triggers carcinogenesis.These results provide new insights into the mechanism of hepatocellular carcinogenesis,as well as new perspectives for understanding the molecular mechanism of hepatocellular carcinoma,and new clues for the diagnosis and treatment of hepatocellular carcinoma.