Develop Prediction Model Of Triple Negative Breast Cancer(TNBC) Based On Machine Learning

In this paper,we provide a high accuracy prediction model for Triple Negative Breast Cancer(TNBC)and explore some possible biomarkers of Triple Negative Breast Cancer.Firstly,we used meta-analysis to validate the existing biomarkers of TNBC(EGFR,VEGF,p53,AR),we search studies on TNBC and biomarkers of TNBC in NCBI,Medline,CNKI,etc.After the literature screening,sorting,data extraction,analysis,we pay attention on biomarkers of TNBC、which has been published in the papers.Secondly,we want to find some possible,high specific biomarkers of TNBC.Differential gene analysis was used to identify genes with significant differences.But,there are still a lot of errors because of the significant difference.That is to say,the high expression or low expression of simple biomarkers can’t diagnose everyone.Finally,we use Machine Learning algorithm to develop a prediction model of TNBC and identify biomarkers of TNBC,such as decision tree and SVM-rfe.Using this method,we can break through the shortcomings of traditional diagnostic methods and achieve a higher accuracy prediction.In the future,we also hope develop a portable detection device,combining with our prediction model,such as mobile terminal device(mobile phone,ipad,etc),to realize the integration of detection,diagnosis and monitoring.