Circulating miRNA Expression Profiling in Breast Cancer Molecular Subtypes: Applying Machine Learning Analysis in Bioinformatics
Alexandra Triantafyllou, NIKOLAOS DOVROLIS, Eleni Zografos, Charalampos Theodoropoulos, George Zografos, Nikolaos V. Michalopoulos, Maria Gazouli
Abstract
BACKGROUND/AIM: Breast cancer is a leading worldwide cause of female cancer-related morbidity and mortality. Since molecular characteristics increasingly guide disease management, demystifying breast tumor miRNA signature emerges as an essential step toward personalized care. This study aimed to investigate the variations in circulating miRNA expression profiles between breast cancer subtypes and healthy controls and to identify relevant target genes and molecular functions. MATERIALS AND METHODS: MiRNA expression was tested by miScript™ miRNA PCR Array Human Cancer Pathway Finder kit, and subsequently, a machine learning approach was applied for miRNA profiling of the various breast cancer molecular subtypes. RESULTS: Serum samples from patients with primary breast cancer (n=66) and healthy controls (n=16) were analyzed. MiR-21 was the single common molecule among all breast cancer subtypes. Furthermore, several miRNAs were found to be differentially expressed explicitly in the different subtypes; luminal A (miR-23b, miR-142, miR-29a, miR-181d, miR-16, miR-29b, miR-155, miR-181c), luminal B (miR-148a, let-7d, miR-92a, miR-34c, let-7b, miR-15a), HER2+ (miR-125b, miR-134, miR-98, miR-143, miR-138, miR-135b) and triple negative breast cancer (miR-17, miR-150, miR-210, miR-372, let-7f, miR-191, miR-133b, miR-146b, miR-7). Finally, miRNA-associated target genes and molecular functions were identified. CONCLUSION: Applying a machine learning approach to delineate miRNA signatures of various breast cancer molecular subtypes allows further understanding of molecular disease characteristics that can prove clinically relevant.