More than 30% of drugs exert their effects by modulating the activity of G protein-coupled receptors (GPCRs) as agonists or antagonists. To understand the working mechanism of drugs and design new ones, the 3D structural information on therapeutic target proteins is crucial. However, every previous approach to determine the structures of GPCRs requires time-consuming experimental screenings of the expression construct for each target. This process significantly hinders high-throughput structural analysis of GPCRs. Moreover, there is no universal strategy for cryo-EM analysis of GPCRs in both agonist- and antagonist-bound forms.
Here, I present a new method for rapid cryo-EM structure determination of GPCRs, called NOAH (NOvel AI-assisted High-throughput construct screening for structural analysis). NOAH is a program that automatically generates the expression constructs of soluble protein-fused GPCRs suitable for cryo-EM analysis. By employing the NOAH pipeline, we can skip the process of experimental screening, saving a significant amount of time and resources on the project. As a proof-of-concept experiment, I applied this method to three GPCRs and determined not only the antagonist-bound structures but also an agonist-bound structure, demonstrating NOAH‘s potential to facilitate GPCR structural biology and drug discovery.