Identification of hsa-mir-۱۸۳ and hsa-mir-۱۳۹ as two potencial candidates for therapeutic aims in breast Cancer through bioinformatics and machine learning approaches

Breast cancer is the most prevalent cancer among women and the second most common cancer worldwide, making it a leading cause of cancer-related mortality in women. Therefore, it has remained a significant concern in women's health field. Recent advancements in next-generation sequencing have revolutionized bioinformatics and targeted therapies, suggesting the importance of identifying new molecular targets for treatment and prognosis, as well as deepening our understanding of the disease's biological mechanisms. MicroRNAs (miRNAs), a class of small non-coding RNAs, have been proven to act as both oncogenes and tumor suppressors by regulating multiple target genes which make them potential candidates for therapeutic targeting. In this study, we analyzed miRNA-seq data from breast cancer tumors and matched normal tissues from the BRCA dataset of TCGA to identify miRNAs that are differentially expressed in breast cancer and we indicated the presence of ۹۰ differentially expressed miRNAs in this malignancy. By applying logistic Lasso regression, a machine learning algorithm used for feature selection on these ۹۰ differentially expressed miRNAs, we identified two miRNAs—hsa-mir-۱۸۳ and hsa-mir-۱۳۹—that appear to be highly correlated with breast cancer. To further explore their biological significance, we constructed a protein-protein interaction network based on their gene targets, resulting in ۴۳۴ nodes and ۸۴۶ edges. Using the MCODE and Centiscape ۲.۲ plugins in Cytoscape, we recognized ۱۲ hub genes (CTNB۱, ADCY۱, LRP۲, ROCK۲, SPO۱۱, LCK, MSL۳, EBF۱, IDH۲, HNRNPC, PSEN۲, RCN۲) that might play crucial roles in breast cancer progression. Gene Ontology (GO) analysis proposed that the target genes of these two miRNAs could be involved in critical biological processes such as endothelial cell differentiation, endothelium development, and the cAMP metabolic process. Reactome pathway enrichment analysis further indicated that these genes might be part of key signaling pathways, including Signaling by Receptor Tyrosine Kinases, MAP kinase activation, and Interleukin-۱۷ signaling. Overall, our study suggests the potential of combining bioinformatics methods with machine learning to identify two miRNAs, hsa-mir-۱۸۳ and hsa-mir-۱۳۹, as Feasible biomarker candidate for breast cancer, which may serve as valuable therapeutic targets and prognostic indicators, and These findings could be further validated in later studies through other methodologies.

کلیدواژه ها:

breast cancer ، differentially expressed microRNAs ، Lasso regression ، Biomarker ، protein-protein interaction