Machine Learning Guided Evolution of a Highly Selective and Sensitive Serotonin Sensor

This research paper explores the use of machine learning to direct the evolution of a serotonin (5-HT) sensor with enhanced selectivity and sensitivity. The authors utilize a luciferase-based system to establish the machine learning approach. Before pursuing a more ambitious goal of completely transforming a choline binding protein into a 5-HT specific sensor, they first established the machine learning approach on a modified ligand/binding pocket. This strategy allows for a better understanding and control of the parameters influencing the performance of the sensor.