DimReduction - Interactive Graphic Environment for Dimensionality Reduction

TL;DR

DimReduction offers an interactive, open-source graphical platform for applying, testing, and comparing various feature selection algorithms specifically tailored for bioinformatics applications like genomics and proteomics.

Contribution

This work introduces a multiplatform graphical environment that facilitates the application and comparison of multiple feature selection methods in bioinformatics.

Findings

Supports various feature selection algorithms

Enables comparison of methods in bioinformatics contexts

Open-source and multiplatform design

Abstract

Feature selection is a pattern recognition approach to choose important variables according to some criteria to distinguish or explain certain phenomena. There are many genomic and proteomic applications which rely on feature selection to answer questions such as: selecting signature genes which are informative about some biological state, e.g. normal tissues and several types of cancer; or defining a network of prediction or inference among elements such as genes, proteins, external stimuli and other elements of interest. In these applications, a recurrent problem is the lack of samples to perform an adequate estimate of the joint probabilities between element states. A myriad of feature selection algorithms and criterion functions are proposed, although it is difficult to point the best solution in general. The intent of this work is to provide an open-source multiplataform graphical environment to apply, test and compare many feature selection approaches suitable to be used in bioinformatics problems.