Building an intelligent integrated method of gene selection for facioscapulohumeral muscular dystrophy diagnosis

Abstract

These days, the genetic testing of neuromuscular diseases is an active area of research. The microarray technology is playing a vital role in analysing the whole genome simultaneously. In microarrays, due to the presence of large number of genes and a small number of samples, it is very difficult to find the disease specific gene subset. So, here gene selection for classification plays a major role in the genetic testing of the disease. In this paper, the gene selection is performed by deploying genetic algorithm with different number of genes wherein the fitness function is evaluated using three different classifiers namely linear discriminant analysis, quadratic discriminant analysis and k-nearest neighbour one-at-a-time for classification of facioscapulohumeral muscular dystrophy. The comparative analysis of the same is also shown in this paper. A facioscapulohumeral muscular dystrophy gene dataset consisting of 50 samples and 33,297 genes is used to evaluate the performance of integrated algorithms. The result shows that the integration of genetic algorithm with k-nearest neighbour is found to be the best for gene selection and diagnosis of facioscapulohumeral muscular dystrophy.