Résultats du processus de synthèse de l'article arXiv : 2201.05408v1

Résumé : MicroRNAs play an indispensable role in numerous biological processes ranging from organismic development to tumor progression.In oncology,these microRNAs constitute a fundamental regulation role in the pathology of cancer that provides the basis for probing into the influences on clinical features through transcriptome data. Previous work focused on machine learning (ML) for searching biomarkers in different cancer databases, but the functions of these biomarkers are fully not clear. Taking lung cancer as a prototype case of study. Through integrating clinical information into the transcripts expression data, we systematically analyzed the effect of microRNA on diagnostic and prognostic factors at deteriorative lung adenocarcinoma (LUAD). After dimension reduction, unsupervised hierarchical clustering was used to find the diagnostic factors which represent the unique expression patterns of microRNA at various patient's stages. In addition, we developed a classification framework, Light Gradient Boosting Machine (LightGBM) and SHAPley Additive explanation (SHAP) algorithm, to screen out the prognostic factors. Enrichment analyses show that the diagnostic and prognostic factors are not only enriched in cancer-related athways, but also involved in many vital cellular signaling transduction and immune responses. These key microRNAs also impact the survival risk of LUAD patients at all (or a specific) stage(s) and some of them target some important Transcription Factors (TF).The key finding is that five microRNAs (hsa-mir-196b, hsa-mir-31, hsa-mir-891a, hsa-mir-34c, and hsa-mir-653) can then serve as not only potential diagnostic factors but also prognostic tools in the monitoring of lung cancer.