Gene Promoter Identification Using Machine Learning

Gene expression is the process composed of two steps i.e. transcription and translation. Transcription involves turning DNA into RNA and translation involves turning RNA into protein. It is a known fact that protein is necessary for proper functioning of body and formation of new cells. So, gene expression is the essential process and this biological process of gene expression begins from promoter sites. By the action of enzyme RNA polymerase on promoter site, the process of transcription starts. In order to understand the whole mechanism of DNA sequences, it is crucial to identify the promoter sites. The paper concentrates on the classification of DNA sequences of Escherichia coli bacteria as promoter and non-promoter. Some classification techniques applied on the dataset. In order to train the model, Convolutional Neural Network, Logistic Regression, Long Short-term Memory and Random Forest Classifier are used. Scikit-learn tool is introduced for the simplification of this binary classification problem. A comparative analysis of used classification techniques is described to understand which technique is more efficient for classifying DNA sequences into promoter and non-promoter.