A MICROFLUIDIC AND MACHINE LEARNING-ENABLED SMARTLAB FOR AUTONOMOUS REMOTE EXECUTION AND ITERATION OF MULTISCALE LIVE CELL ASSAYS FOR DRUG DISCOVERY

临床前药物发现化验的发展依赖于科学家的在场来开发、启动和监测化验。执行这些活动的中断和错误导致重新开始实验的额外时间和费用，以及先前花费的时间和试剂的浪费。该项目设想的创新的SmartLab框架通过实现远程、自主的启动、执行和迭代，解决了临床前基于细胞的分析中中断和错误的重大且代价高昂的风险。 SmartLab框架的优势包括：(1)减少人为错误的风险；(2)当亲自实验室操作中断时，不受阻碍地继续进行实验；(3)通过自适应实验反馈实现实验效率的最大化。总而言之，这些优势将极大地提高用于药物发现的临床前分析框架的稳健性和效率，从而使人类健康受益。

Preclinical drug discovery assay development depends on the presence of scientists to develop, initiate and monitor assays. Interruptions and errors in the execution of these activities incur extra time and expense to restart experiments and the waste of previously expended time and reagents. The innovative SmartLab framework envisioned by this project addresses the significant, costly risks of interruptions and errors in preclinical cell-based assays by enabling their remote, autonomous initiation, execution and iteration. Advantages of the SmartLab framework include (1) reduction in risk of human error; (2) unimpeded continuation of experiments when in-person lab operations are interrupted and (3) maximized experimental efficiency through adaptive experimental feedback. Collectively these advantages will benefit human health by dramatically improving the robustness and efficiency of preclinical assay frameworks used for drug discovery.