Neural-Network Scoring Functions Identify Structurally Novel Estrogen-Receptor Ligands.

Area of Science:

• Medicinal Chemistry
• Computational Biology
• Artificial Intelligence in Drug Discovery

Background:

• Drug development is costly and time-consuming, with high investments required for bringing new therapies to market.
• Innovations that increase the efficiency of drug discovery can accelerate the delivery of treatments to patients.
• Virtual screening (in silico testing) aids in prioritizing compounds for experimental evaluation.

Purpose of the Study:

• To demonstrate the efficacy of neural networks in identifying novel small molecules that bind to drug targets.
• To improve the accuracy and efficiency of virtual screening processes.
• To reduce the number of compounds tested in early-stage drug lead identification.

Main Methods:

• Utilized neural networks for in silico virtual screening.
• Focused on the estrogen receptor as a model drug target.
• Experimentally determined the binding affinity (Ki values) of identified compounds.

Main Results:

• Identified 39 novel estrogen receptor ligands using neural network-based virtual screening.
• Experimentally validated the binding of these novel ligands.
• Achieved a range of experimentally determined Ki values from 460 nM to 20 μM.

Conclusions:

• Neural networks are effective tools for identifying structurally novel small molecules with drug target binding capabilities.
• This AI-driven approach enhances the efficiency of lead identification in drug discovery.
• The study presents novel estrogen receptor ligands for potential therapeutic development.