DNA encoding with entropy based numerical mapping technique for phylogenetic analysis

In order to comprehend the relationship between species and their evolution process, phylogeny analysis is generally applied. DNA sequences are an essential part of that analysis since DNA gives valuable information on each species. Yet, developing a phylogenetic tree requires time and effort which affect both researchers and systems performance in a negative way. In that study, we developed a machine learning method to determine which DNA sequence belongs to which class of species without performing phylogenetic analysis. Firstly, we obtained some data from NCBI datasets and in that study DNA sequences were considered as ND1 (Dehydrogenase Subunit 1) which are obtained from the Mitochondrial DNA sequences of each species. After the data collection process, DNA sequences were converted to numerical values by applying the entropy-based mapping method. In the last part of the study, MLPNN was applied to determine which DNA sequences belong to which class. Results showed that our proposed method gave a better performance which discriminates the species with 96.67% accuracy.