Tashkent and Pushchino Team Uses AI to Design Short Peptides That Selectively Activate AMPA or NMDA Glutamate Receptors in Rat Neurons
Researchers from Uzbekistan's Centre for Advanced Technologies and the Russian Academy of Sciences' Institute of Cell Biophysics combined a generative neural network, AlphaFold2, and molecular dynamics to design short peptides aimed at the AMPA and NMDA glutamate receptors, which carry much of the signaling between neurons and are targets for neurological drugs. Synthesized candidates caused concentration-dependent calcium influx in rat cortical and hippocampal cultures that the blockers CNQX and MK-801 abolished, and peptides built for one receptor did not act on the other. The responses were weaker than those to the standard agonists AMPA and NMDA, the PNAS Nexus paper reports.