Introduction

GeneSelection is a website that introduces current research in the area of feature selection from microarray datasets. This site summarizes most popular and widely used feature selection methods and provides a platform to run different methods with your own microarray datasets.

Currently, GeneSelection collects feature selection methods developed by different researchers. They are: mRMR(minimum Redundancy Maximum Relevance),RankGene which includs Information gain ranking, Twoing rule ranking, Gini index ranking, Sum minority ranking, Max minority ranking,and Sum of variances ranking, t-statistic ranking selection, One dimensional SVM ranking selection, Clustering based method and some other methods.

To learn more and run the program of each method, please click the corresponding links in the side menu. To upload your datasets, please follow the file format described below. You may also select the listed datasets in each method.

All the programs are reserved to non-profit and academic use.

File Format

To upload your own datasets through the GeneSelection site, you need to save your files as standard CSV format. Each row contains the class label and all the values of one sample.

<class_label>,<value1>,<value2>,<value3>,...

The first row must be the feature names, and the first column must be the classes for samples. All the expression values are of float type and class labels are of integer type. You may download a sample dataset below.

Sample dataset: Arrhythmia data