Identification Of Tissue-specific MicroRNAs And Analysis Of Their Regulatory Networks

MicroRNAs(miRNAs)are a class of 18 to 22 nt non-coding RNAs,whose expression disorders cause changes in cell function and biological processes.Understanding the expression patterns of tissue-specific genes is particularly important for elucidating the molecular mechanisms of transcriptional regulation and cellular function.The miRNA can be targeted to the 3'-UTR region as a tissue-specific negative regulatory element,and the gene is degraded and inhibited by the transcription factor(Transcription Factor,TF).Previous studies have found that miRNAs have significant tissue specificity and play important roles in specific tissues.However,there are no reports on the tissue specificity and regulatory networks of miRNAs based on TCGA clinical samples for the identification and analysis of miRNAs.The amount of the miRNA expression in this research by integrating TCGA data,in the adrenal glands,bile duct,bladder,blood,brain,breast,cervix,colorectal,esophagus,head and neck,kidney,liver,lung,pancreas,prostate,skin,stomach,thymus,thyroid,uterus,a total of 20 group match in normal and cancer tissues identified 930 effective expression of microRNAs in the sample.Three types of mirnas were screened by TSI:tissue specific miRNAs(330 in normal tissue,253 in cancer tissue),widely expressed miRNAs(67 in normal tissue,106 in cancer tissue),and other miRNAs(297 in normal tissue,280 in cancer tissue).Tissue-specific miRNA shared by normal tissues and cancer tissues showed significant differences in all tissues,and their TSI was proportional to the coefficient of variation,indicating that tissue specific miRNA had greater impact on tissues when miRNA expression was misaligned than when miRNA expression was widely expressed.By comparing the miRNA expression levels of normal and cancer tissues,a total of 1,013 miRNAs were significantly up-regulated,and 843 miRNAs were significantly down-regulated.There was a significant difference in the quantity distribution of the two in a single tissue,and tissue specific miRNAs tended to be up-regulated.By analyzing the GO function enrichment of up-regulated miRNA,it was found that it was associated with 29 cancer biological processes such as cell adhesion,autophagy and anti-autophagy,and immune response,while down-regulated miRNA was significantly associated with 36 cancer biological processes such as cell autophagy,cell migration,cell cycle,cell differentiation,cell growth,and cell proliferation.Enrichment analysis of up-regulated miRNA KEGG signaling pathway revealed that it was associated with 23 cancer signaling pathways such as TNF,p38 MAPK,ATR and Wnt,while down-regulated miRNA was significantly associated with 25 cancer signaling pathways such as RAC1,TNF and p38 MAPK.Integrated the interaction between transcription factors and miRNA,miRNA-target,as well as transcription factors and genes,the feedforward loop is of great significance for the better study of tissue-specific miRNA.Therefore,this research developed an interactive database TSmiRNA(http://lcbb.swjtu.edu.cn/TSmiRNA/),facilitated to search and download the data for user.Currently,the database contains a total of 416 tissue-specific miRNAs from normal and cancer tissues with 20 matched pairs,126 widely expressed miRNAs and 388 other miRNAs.137 transcription factors regulating miRNA,20265 target genes of miRNA,expression level data of miRNA,transcription factors,genes,TSI,coefficient of variation,transcription factor binding sites,and other information were also included in 13 regulation relationships of transcription factor-miRNA-gene in normal and cancer tissues,totaling 295951 and 275927.By integrating the miRNA expression data of normal breast and breast cancer tissues in TCGA,132 differentially expressed miRNAs were identified,and the expression imbalance of differentially expressed mirnas was evaluated using the coefficient of variation.Fltrated the differentially expressed miRNA with the coefficient of variation ranking the top 15%,a total of 19 breast cancer-causing miRNAs were obtained,including 5 tissue-specific miRNAs.Construction of miRNA regulatory network for breast cancer-causing.including 5 miRNAs,8 transcription factors and 130 genes,a total of 262 cancer regulatory networks.Degree centrality and average clustering coefficient were used to analyze and analyze the network architecture,and it was found that mirnas regulate each other through common transcription factors and target genes.GO analysis found that these 5 miRNA target genes were related to the biological processes of tumor,such as cell cycle,cell differentiation,cell growth and other post-transcriptional regulation.KEGG analysis showed that these 5 miRNAs were highly correlated with cancer signaling pathways such as the p53 signaling pathway.The results of survival analysis showed that the expression of hsa-mir-144 was inversely proportional to the survival rate of hsa-mir-133a-2,suggesting a significant correlation with survival of breast cancer patients.