Deciphering the c-MYC miRNA Regulatory Network through Machine Learning

Introduction: The oncogene c-MYC plays a pivotal role in cancer development andprogression[۱], yet its intricate control network, particularly involving microRNAs (miRNAs), remainselusive[۲]. This study delves into c-MYC's potential to encode miRNA regulatory information usingmachine learning, aiming to decipher its miRNA regulatory landscape.Materials and Methods: We compiled diverse datasets encompassing c-MYC expression, miRNAexpression, and predicted miRNA target sites. Using feature engineering techniques, we extractedsequence and positional features surrounding potential miRNA target sites within the c-MYC gene.Subsequently, we employed different machine learning algorithms, including random forests andsupport vector machines, to classify true miRNA target sites based on these features. We evaluatedmodel performance using cross-validation and compared results across algorithms.Results: Our models achieved high accuracy in identifying true miRNA target sites within the c-MYCgene, exceeding ۸۵% in most cases. By analyzing the learned feature weights, we uncovered keysequence and positional motifs crucial for miRNA recognition and binding. Interestingly, we identifiedclusters of enriched motifs within the c-MYC gene, suggesting the presence of miRNA regulatoryhotspots. Furthermore, we mapped these hotspots to specific functional domains within c-MYC,revealing a potential link between miRNA regulation and c-MYC function.Conclusions: Our study demonstrates the feasibility of using machine learning to decode miRNAregulatory networks encoded within genes. By applying this approach to c-MYC, we unveiled acomplex miRNA regulatory landscape with hotspots likely linked to specific c-MYC functions. Thisunderstanding paves the way for further investigations into miRNA-mediated control of c-MYC incancer. Future research will focus on validating predicted target sites experimentally and elucidatingthe functional consequences of this newly discovered miRNA regulatory network in cancer biology.

کلیدواژه ها:

c-MYC ، microRNA ، miRNA regulatory network ، machine learning ، target site prediction ، cancer