This comprehensive thesis investigates a range of computational approaches in peroxisomal research. It focuses on utilizing deep learning-based protein sequence embeddings to predict sub-peroxisomal protein localization, peroxisomal protein functions and their interactions. Additionally, it addresses the semantic interpretation of bioassays through Natural Language Processing (NLP). The thesis also encompasses bioinformatic training initiatives, promoting knowledge dissemination. Furthermore, it explores strategies for advancing bioinformatics education, contributing critical thinking skills. By integrating computational methods, predictive tools, NLP, and education projects, this thesis provides a multifaceted contribution to peroxisomal research, enhancing our understanding of peroxisomal functions and their broader implications.