Efficacy and accuracy responses of DNA mini-barcodes in species identification under a supervised machine learning approach

Abstract

Specific gene regions in DNA, like COI (Cytochrome c oxidase I) in case of animals, have been defined as DNA barcode and many studies proved that it can be used as an identifier to distinguish species. The standard length of a DNA barcode is approximately 650 bp. But because of the challenges in sequencing technologies or unavailability of high-quality genomic DNA, it is not always possible to get the full barcode sequence of an organism. As a result, recent studies suggest that mini-barcodes can provide a good contribution in the species identification process. Among various methods proposed for the identification task, supervised machine learning methods have been shown effective. In this study, we have analyzed the effect of different barcode lengths on species identification from the perspective of supervised machine learning and suggested a general approximation of required length of mini-barcode in this regard. We have implemented a Naïve Bayes classifier as our model and implied the effectiveness of mini-barcode by demonstrating the accuracy responses varying the length of DNA barcode sequences.