Study On The Invasiveness Of Breast Cancer Based On DNA Methylation

Tumor metastasis is the key to its high mortality.More and more scholars began to pay attention to the study of tumor metastasis,and achieved satisfactory results in this field.Breast cancer is the most common malignant tumor in women.Its high mortality rate is mainly due to metastasis and recurrence.Invasive and non-invasive breast cancer patients need different treatment methods,so there is an urgent need for predictive tools to guide clinical decision-making,also to avoid over treatment of non-invasive breast cancer and under treatment of invasive cases.With the advent of the era of sequencing,we can study tumor metastasis at the molecular level,which is of great significance to understand the molecular mechanism of tumor metastasis,identify diagnostic markers and therapeutic targets,and guide clinical decision-making.In this study,the expression profile of Illumina Infinium 450 k methylation was used to study the invasiveness of breast cancer.It mainly includes the following aspects:Firstly,the sample set was divided according to the methylation distance of whole genome DNA;Then,based on the 450 k methylation data of the tumor,the purity of the tumor is evaluated and corrected.Mepurity is an algorithm based on the β mixed model.It only needs the DNA methylation data of the tumor sample to calculate the purity of the tumor,but does not need the matched normal sample.DESeq2 is a Differential analysis method,which allows users to provide tumor purity,thus reducing the impact of tumor impure on subsequent analysis.Two differential methylation analysis methods were used to identify the specific Cp G site and get the intersection of the results.In order to reduce the false-positive of differential sites screened based on statistical test,two algorithms were used to carry out the difference analysis.Then,reduce the dimension of the set of differential sites,based on the results of four dimensionality reduction algorithms,use random forest algorithm to construct methylation classifiers to classify primary breast cancer.Finally,the performance of the classifier is verified and evaluated from multiple perspectives by using the hm450 DNA methylation data of breast cancer(BRCA)and clinical data from the tumor Genome Atlas(TCGA)database,including basic model evaluation indicators,clinical factor enrichment analysis based on hypergeometric distribution,hierarchical clustering based on the best feature collection,and literature verification of tumor metastasis related genes.In summary,this study shows the potential of DNA methylation as a biomarker for predicting tumor invasiveness and providing new information for metastatic cancer research.In addition,a website based on differential methylation analysis,dimensionality reduction and classification algorithm is developed in order to facilitate researchers to study and predict the invasiveness of breast cancer.