A Survey on Deep Learning Approaches Used in Genomics

Abstract

Deep learning (DL) methods have shown remarkable success in addressing various problems across different domains. Classifying DNA sequences presents a formidable challenge in the field of bioinformatics. This review delves into various technologies centered around Alignment methods and Deep Learning for the purpose of classification. The aim is to achieve accurate and scalable predictions for DNA sequence classification. DL methods have proven effective in overcoming the primary challenges faced during the training process. The paper delves into previous classification methods like alignment methods and highlights their limitations. Subsequently, we delve into the application of deep learning, specifically using CNN and RNN models, for DNA sequence classification. We evaluate their respective accuracies and discuss the differences and drawbacks associated with these methods.