Topic Based On The Probability Model Of Functional Micrornas - Mrna Control Module Identification

MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression at the post-transcriptional level through completely or partially complementary base binding to their target mRNAs, leading to the degradation or translation inhibition of their target mRNAs. miRNAs play important roles in multiply biological and metabolic processes, including developmental timing, cell differentiation, proliferation, growth, migration, apoptosis, stress responses and cancer, thus, miRNAs have been considered as one of the key regulators in complicated networks of functional gene regulatory modules. Although much work has been done to elucidate the regulatory mechanism of miRNAs by associating miRNAs with mRNAs, their precise functions are still largely unknown. In this paper, we proposed a probabilistic topic model to infer regulatory networks of miRNAs and their targets mRNAs for specific biological conditions at the post-transcriptional level, so-called functional miRNA-mRNA regulatory modules (FMRMs).The FMRMs are produced by combining heterogeneous data sets of epithelial to mesenchymal transition (EMT), including expression profiles of miRNAs and mRNAs, and putative target binding information. The probabilistic topic model used in this paper can effectively capture the relationship between miRNAs and mRNAs in specific cellular conditions. Furthermore, the proposed method finds differentially expressed miRNAs and mRNAs, which are highly relevant to EMT. This approach also identifies negatively and positively correlated miRNA-mRNA pairs which are associated with epithelial, mesenchymal, and other conditions in EMT data sets, respectively. We computed the number of the negative and positive expression miRNA-mRNA pairs, which demonstrates that most miRNA-mRNA pairs are negatively correlated in EMT data sets. The significant results of functional miRNA-mRNA regulatory modules were validated with IPA (Ingenuity Pathway Analysis) software. Finally, compared with K-means clustering approach, probabilistic topic model is more efficient in mining functional miRNA-mRNA regulatory modules.This paper combines heterogeneous data sets of epithelial to mesenchymal transition (EMT) to identify functional miRNA-mRNA regulatory modules in sequence and gene expression level. The results on EMT data sets show that the inferred FMRMs can construct the biological chain of’miRNAâ†'mRNAâ†'condition’at the post-transcriptional level and give new insights into biological process, functional network of many diseases and miRNA targets therapy.