Classification of Eukaryotic Splice-junction Genetic Sequences Using Averaged One-dependence Estimators with Subsumption Resolution

Abstract

DNA is the building block of life, which contains encoded genetic instructions for building living organisms. Because of the fact that proteins are constructed in accordance with the genetic instructions encoded in DNAs, errors in RNA synthesis and translation into proteins can cause genetic disorders. Therefore, understanding and recognizing genetic sequences is one step towards the treatment of these genetic disorders. Since the discovery of DNA, there has been a growing interest in the problem of genetic sequence recognition, motivated by its enormous potential to cure a wide range of genetic disorders. The completion of the human genome project in the last decade has generated a strong demand in computational analysis techniques in order to fully exploit the acquired human genome database. This paper describes a state-of-the-art machine learning based approach called averaged one-dependence estimators with subsumption resolution to tackle the problem of recognizing an important class of genetic sequences known as eukaryotic splice junctions. To lower the computational complexity and to increase the generalization capability of the system, we employ a genetic algorithm to select relevant nucleotides that are directly responsible for splice-junction recognition. We carried out experiments on a dataset extracted from the biological literature. This proposed system has achieved an accuracy of 96.68% in classifying splice-junction genetic sequences. The experimental results demonstrate the efficacy of our framework and encourage us to apply the framework on other types of genetic sequences.