EUGENe: Revolutionizing Genomic Sequence Analysis with Deep Learning
EUGENe is a new toolkit developed for the analysis of sequence-based datasets in genomics research. It aims to streamline the end-to-end execution of deep-learning workflows, making it easier to design, implement, validate, and interpret deep-learning solutions in regulatory genomics. EUGENe offers functionality for data extraction, transformation, loading, model instantiation, initialization, training, evaluation, and interpretation. It has been successfully applied to tasks such as predicting plant promoter activity, RNA binding protein specificity, and DNA protein binding. The toolkit provides visualization tools to identify key sequence features used by models to make predictions, such as core promoter elements and transcription factor binding motifs. EUGENe aims to promote the adoption of FAIR (findable, accessible, interoperable, and reusable) data and software principles in genomics research.