Revealing miRNA Biomarkers for Alzheimer s Disease using NGS

This main aim of this project is to find a set of miRNA biomarkers for Alzheimer’s Disease using Next Generation Sequencing Data. A dataset obtained from NCBI database under the access number GSE46579 is used. To achieve the goal of finding the best set of differentially expressed miRNAs, we followed a step wise procedure including five main steps as data collection, preprocessing, statistical analysis, feature selection and classification. Under the data collection stage, initial data set was chosen. We used Galaxy platform for preprocessing. The statistical analysis included significance value (pValue) analysis, fold change analysis followed by the area under receiver operating curve (AUC) analysis. For the features reduced from the statistical methods, feature selection was done under three methods as PCA, Random Forest and correlation coefficient analysis. We used classification algorithms for presenting the possibilty of miRNAs in making better predictions. In this project, we used python programming language for both statistical analysis and machine learning analysis of the data.