Promoter Prediction in Helicobacter pylori Strain 26695 by Using a Deep Neural Network

To initiate gene transcription, RNA polymerase binds to the upstream genomic region of the gene called the promoter. As the most important component of the genome, gene expression is regulated by promoter identification. The present work employed a deep learning technique to predict promoter sequences in Helicobacter pylori (H. pylori) strain 26695. Here, we used evolutionary scale modeling (ESM) and one-hot encoding to encode the promoter sequence, and then these encoded features were given as an input to the convolutional neural network (CNN) based classifier by utilizing tenfold cross-validation (CV). The results of the anticipated model showed reasonable performance, with an accuracy of 88.71%.