A Fast Multiple Attractor Cellular Automata with Modified Clonal Classifier for Splicing Site Prediction in Human Genome

TL;DR

This paper introduces a fast cellular automata-based classifier with a modified clonal algorithm for accurately predicting splicing sites in human genome DNA sequences, demonstrating promising results compared to existing methods.

Contribution

It proposes a novel cellular automata model combined with a modified clonal algorithm for improved splicing site prediction in bioinformatics.

Findings

Achieved high accuracy on Irvine Primate Splice Junction Database.

Outperformed existing methods like NNspIICE, GENIO, HSPL, and SPIICE VIEW.

Demonstrated efficiency and potential for biological data analysis.

Abstract

Bioinformatics encompass storing, analyzing and interpreting the biological data. Most of the challenges for Machine Learning methods like Cellular Automata is to furnish the functional information with the corresponding biological sequences. In eukaryotes DNA is divided into introns and exons. The introns will be removed to make the coding region by a process called splicing. By indentifying a splice site we can easily specify the DNA sequence category (Donor/Accepter/Neither).Splicing sites play an important role in understanding the genes. A class of CA which can handle fuzzy logic is employed with modified clonal algorithm is proposed to identify the splicing site. This classifier is tested with Irvine Primate Splice Junction Database. It is compared with NNspIICE, GENIO, HSPL and SPIICE VIEW. The reported accuracy and efficiency of prediction is quite promising.