EnsembleGASVR: a novel ensemble method for classifying missense single nucleotide polymorphisms

Abstract

Datasets and codes are freely available on the Web at http://prlab.ceid.upatras.gr/EnsembleGASVR/dataset-codes.zip. All the required information about the article is available through http://prlab.ceid.upatras.gr/EnsembleGASVR/site.html.

On the Computational Prediction of miRNA Promoters

“…Elitism is used to force the best solution of each population to be selected at least once in the next generation. This selection mechanism has been suggested in [27].…”

“…The proposed method is an embedded classification method that combines an adaptive GA with a nu-SVR classifier. It is inspired by EnsembleGASVR, a method suggested in [27] for classifying missense single nucleotide polymorphisms. In principle, SVR classifiers present high classification performance and low complexity.…”

“…The number of selected features follows next and the number of support vectors is the least significant objective. To avoid over-fitting problems, we did not attempt to optimize these values, as suggested in [27]. The weights of the goals have been set so as to achieve high classification performance and simultaneously generate a simple and effective model.…”

“…The crossover rate is constant and set to 0.9, in order to leave some part of the population to survive unchanged to the next generation. This property is essential when good solutions emerge in early stages of the algorithm, as proposed in [27].…”

Motif signatures of transcribed enhancers

“…In phase (1), to identify candidate combinations of motifs TELS uses filtering feature selection (FS) techniques. The FS problem in bioinformatics is very well studied [25][26][27][28][29][30][31][32][33] and it is well documented that FS is a strongly 'data-dependent' process. TELS first ranks the 346 individual variables using the Gini-index based FS.…”