Fully autonomous robotic wet lab for stem cell culture media optimization

PROJECT SUMMARY Hematopoietic stem cells (HSCs) are perhaps the most important and immediately actionable cells for cell rejuvenation therapies, but the difficulty of growing them in the lab limits their potential. To address this problem, we are using a fully autonomous, ML-driven wet lab to optimize HSC culture media. Our system, which we refer to as a “machine biologist”, has two major components: the “body”, consisting of a robotic wet lab capable of performing HSC culture experiments, and the “mind”, consisting of an machine learned (ML) computer vision model for analyzing our microscopy results and a Bayesian optimizer to perform experimental design. This combination will allow our system to operate fully autonomously, and proceed orders-of-magnitude faster than is possible in a conventional biology lab. Our work has the potential for transformative impact on both the way cell biology research is conducted as well as HSC-based cell therapies.