An efficient deep learning-based strategy to screen inhibitors for GluN1/GluN3A receptor

Area of Science:

• Neuroscience
• Pharmacology
• Computational Biology

Background:

• The GluN1/GluN3A receptor, an excitatory glycine receptor in the central nervous system, impacts emotional regulation and presents a therapeutic target for neuropsychiatric disorders.
• Current pharmacological research on GluN1/GluN3A receptors is limited, and traditional high-throughput screening methods are inefficient for large compound libraries.

Purpose of the Study:

• To design a deep learning-based strategy for efficient and accurate identification of GluN1/GluN3A inhibitors.
• To overcome the limitations of traditional screening methods in drug discovery for ion channels.

Main Methods:

• Developed a sequence-based scoring function to rapidly screen an 18 million compound library, reducing candidates to ~10^5.
• Utilized complex-based scoring functions (IGModel and RTMScore) for precise scoring and ranking of remaining candidates.
• Confirmed an active molecule using whole-cell voltage-clamp electrophysiology.

Main Results:

• Successfully identified a potent GluN1/GluN3A inhibitor with an IC50 of 2.87 ± 0.80 μM.
• The deep learning strategy significantly improved the efficiency and accuracy of high-throughput screening.
• Demonstrated the feasibility of integrating deep learning into ion channel drug discovery.

Conclusions:

• The developed deep learning strategy offers a rapid and precise approach for identifying ion channel modulators.
• This study provides a new paradigm for drug discovery targeting novel receptors like GluN1/GluN3A.
• The findings pave the way for developing therapeutics for neuropsychiatric disorders.