An Evaluation of DNA Barcoding Using Genetic Programming-Based Process

The DNA barcoding is a promising technique for identifications of biological species based on a relatively short sequence of COI gene. A research area to improve the DNA barcoding is to study the classification techniques that utilize common properties of DNA and amino acid sequences such as variable lengths of gene sequences, and the comparison of different reference genes. In this study, we evaluate a classification model for DNA barcoding induced by genetic programming. The proposed method can be adapted for both DNA and amino acid sequences. The performance is evaluated by representing the two types of sequences and one based on their properties. The proposed method evaluates common significant sites on the reference genes which are useful to differentiate between species.