A Review of Soft Computing Techniques for Gene Prediction

In the past decade, various genomes have been sequenced in both plants and animals. The falling cost of genome sequencing manifests a great impact on the research community with respect to annotation of genomes. Genome annotation helps in understanding the biological functions of the sequences of these genomes. Gene prediction is one of the most important aspects of genome annotation and it is an open research problem in bioinformatics. A large number of techniques for gene prediction have been developed over the past few years. In this paper a theoretical review of soft computing techniques for gene prediction is presented. The problem of gene prediction, along with the issues involved in it, is first described. A brief description of soft computing techniques, before discussing their application to gene prediction, is then provided. In addition, a list of different soft computing techniques for gene prediction is compiled. Finally some limitations of the current research and future research directions are presented. 1. Introduction In the past several years, there has been a virtual explosion of genomic sequence data with numerous of genomes in various stages of sequencing and annotation. As the human genome project came to an end in 2003, all the human chromosomes have been sequenced [1]. In fact, with the number of genomes sequenced numbering over one hundred, it is clear that quick, accurate annotation of these genomes is essential to learning more about biology and the evolutionary relationships between these genomes [2]. However, the pace of genome annotation is not matching the pace of genome sequencing. The experimental annotation of genomes is slow and time consuming. Therefore there is a real need to develop automatic techniques for genome annotation. The first step towards successful genome annotation is gene prediction. Gene prediction is mainly concerned with the identification of protein-coding genes in DNA but may also include the identification of other functional elements of genomic DNA such as RNA genes and regulatory regions. A large number of techniques have been developed for the prediction of protein-coding genes. However the prediction accuracy of these techniques is still far from satisfactory. There are two main problems with the existing protein-coding gene prediction techniques. First, most of the techniques are developed for specific genomes. Second, the gene level accuracy of these techniques is very low. It is obvious that further improvement to protein-coding gene prediction is much needed. An extensive list of existing