Self-driving laboratories for autonomous exploration of protein sequence space

项目总结/摘要 我们建议开发完全自主的“自动驾驶实验室”，以快速设计酶， 在生物医学和生物催化中的应用。我们的方法模仿了蛋白质工程的方法 研究人员与人工智能层，建立蛋白质序列的理解-结构-功能和计划 实验来测试特定的蛋白质设计假设，以及一个机器人系统， 通过合成基因、表达蛋白质和进行生物化学测量来设计蛋白质， 酶活性智能代理和实验自动化之间的无缝集成， 自主设计-测试-学习循环，以理解和优化序列-功能景观。自我 驾驶实验室将通过自动化， 非常低效、耗时和费力的蛋白质工程活动，能够快速周转， 并使研究人员能够集中精力研究重要的下游应用。

PROJECT SUMMARY/ABSTRACT We propose to develop fully autonomous “self-driving laboratories” to rapidly engineer enzymes for broad applications in biomedicine and biocatalysis. Our approach mimics the methodology of a protein engineering researcher with an AI layer that builds an understanding of protein sequence-structure-function and plans experiments to test specific protein design hypotheses, and a robotic system that experimentally tests designed proteins by synthesizing genes, expressing proteins, and performing biochemical measurements of enzyme activity. Seamless integration between the intelligent agent and experimental automation enables fully autonomous design-test-learn cycles to understand and optimize the sequence-function landscape. Self- driving laboratories will revolutionize the fields of biomolecular engineering and synthetic biology by automating highly inefficient, time consuming, and laborious protein engineering campaigns, enabling rapid turnaround, and allowing researchers to focus efforts on important downstream applications.