Estimation of relevant variables on high-dimensional biological patterns using iterated weighted kernel functions

Abstract

Background: The analysis of complex proteomic and genomic profiles involves the idenfication of significant makers within a set of hundreds or even thousands of variables that represent a high-dimensional problem space. The occurrence of noise, redundancy of combinatorial interactions in the profile makes the selection of relevant harder. Methodology/Principal Findings: Here we propose a method to select variables based on estimated relevance to hidden patterns. Our method combines a weighted-kemel discriminant with an iterative stochastic probability estimation algorithm to discovery the relevance distribution over the set of variables. We verified the ability of our method to select predefined relevant variables in synthetic proteome-like data and then assessed its performance on biological high-dimensional problems. Experiments were run on serum proteomic datasets of infectious diseases. The resulting variable subsets achieved classification accuracies of 99% on Human African Trypanosomias, 91% on Tuberculosis , and 91% on Malaria serum proteomic profiles with fewer than 20% of variables selected. Our method scaled-up to dimensionalities of much higher orders of magnitude as shown with gene expression microarray datasets in which we obtained classification accuracies close to 90% with fewer than 1% of the total number of variables. Conclusions: Our method consistently found relevant variables attaining high classification accuracies across synthetic and biological datasets. Notably, it yielded very compact subsets compact to the original number of variables, which should simplify downstream biological experimentation. © 2008 Rojas-Galeano et al.