Annotating genomes with DeepGO protein function prediction tools
Abstract
This chapter explores the evolution of DeepGO, a suite of deep learning-based tools for protein function prediction, in the form of Gene Ontology (GO) terms, and their applications in genome annotation. We provide a comprehensive overview of the different versions of DeepGO, highlighting key advancements introduced by each method. To demonstrate the practical application of these tools, we present a case study on the annotation of a bacterial genome using the latest Deep GO model, DeepGO-SE. We showcase the efficiency and accuracy of DeepGO-SE in predicting protein functions and discuss the model’s parameters. This chapter serves as a guide for researchers looking to enhance their genomic analyses using deep learning-based function prediction methods.