Novel Approach to Predict Promoter Region Based on Short Range Interaction Between DNA Sequences

Abstract

Genomic studies have become one of the useful aspects of Bioinformatics since it provides important information about an organism’s genome once it has been sequenced. Gene finding and promoter predictions are common strategies used in modern Bioinformatics which helps in the provision of an organism’s genomic information. Many works has been carried out on promoter prediction by various scientists and therefore many prediction tools are available. However, there is a high demand for novel prediction tools due to low level of prediction accuracy and sensitivity which are the important features of a good prediction tool. In this paper, we have developed the new algorithm Novel Approach to Promoter Prediction (NAPPR) to predict eukaryotic promoter region using the python programming, which can meet today’s demand to some extent. We have developed the parameters for Singlet (4$^{1}$) to nanoplets (4$^{9}$) in analyzing short range interactions between the four nucleotide bases in DNA sequences. Using this parameters NAPPR tool was developed to predict promoters with high level of Accuracy, Sensitivity and Specificity after comparing it with other known prediction tools. An Accuracy of 74 % and Specificity of 78 % was achieved after testing it on test sequences from the EPD database. The length of DNA sequence used as input has no limit and can therefore be used to predict promoters even in the whole human genome. At the end, it was found out that NAPPR can predict eukaryotic promoter with high level of accuracy and sensitivity.