Protein design is a rapidly evolving field, with machine learning taking an increasingly important role in the development of novel protein biotechnologies. Yet, computer-generated protein designs only become impactful contributions if validated experimentally. We aim to build a fully automated platform for high-throughput synthesis of protein designs and measurement of their function, resulting in high-quality datasets that should be employed in a machine learning directed evolution (MLDE, active learning) cycle to optimize protein designs for activity, expression, and yield. The project is a multi-institutional collaborative effort, is set to start in January 2026, and is funded for two years by the Bavarian Ministry of Economic Affairs.
The successful applicant will hold a Ph.D. in computer science, machine learning, bioinformatics, or related fields. The applicant should have a strong interest and prior experience in developing machine learning methods and/or applying machine learning algorithms for solving problems in protein engineering. Written and oral command of English is essential.