DFA7, a New Method to Distinguish between Intron-Containing and Intronless Genes

Intron-containing and intronless genes have different biological properties and statistical characteristics. Here we propose a new computational method to distinguish between intron-containing and intronless gene sequences. Seven feature parameters alpha, beta, gamma, lambda, theta, phi, and sigma based on detrended fluctuation analysis (DFA) are fully used, and thus we can compute a 7-dimensional feature vector for any given gene sequence to be discriminated. Furthermore, support vector machine (SVM) classifier with Gaussian radial basis kernel function is performed on this feature space to classify the genes into intron-containing and intronless. We investigate the performance of the proposed method in comparison with other state-of-the-art algorithms on biological datasets. The experimental results show that our new method significantly improves the accuracy over those existing techniques.