In this seminar, Sven Hauns will delve into the innovative application of deep learning systems for the annotation of bacterial immune systems. The primary focus will be on the development and calibration of deep learning models capable of accurately classifying immune systems in entire genomes. A crucial aspect of the discussion will be the necessity for well-calibrated models, emphasizing the challenge of discerning and rejecting unrelated proteins. The presentation will address the open space problem in deep learning, shedding light on the inherent complexities and uncertainties faced in this evolving field. Sven will share insights into the prospects for the discovery of novel immune systems, showcasing the potential impact of his research on advancing our understanding of bacterial defense mechanisms. Attendees can expect an exploration of the intersection between deep learning and microbiology, with practical applications in the annotation and classification of bacterial immune systems.