Multiscale Brain Model Predicts Novel Propofol Anesthesia Biomarker Without Training on Clinical Data

A team of researchers has introduced a multiscale mechanistic model of thalamocortical brain architecture to study how propofol, a common anesthetic, reorganizes brain activity from the receptor level up to large-scale circuits. The model incorporates a core-matrix thalamocortical structure and is driven exclusively by GABA-A receptor modulation, with no parameter fitting to anesthesia data. When tested on a standard auditory oddball paradigm, the simulation matched independent macaque electrophysiology datasets without any task-specific training. The same unmodified model also reproduced changes in functional connectivity observed in anesthetized humans, specifically showing selective attenuation of matrix thalamocortical loops relative to core loops. Critically, the simulation predicted a dose-dependent biomarker — elevated residual inter-stimulus cortical activity — that was subsequently confirmed in existing macaque data where it had previously gone unnoticed. The authors argue this 'simulation-first' discovery framework, grounded in mechanistic circuit dynamics rather than statistical population comparisons, represents a generalizable approach for translating receptor-level pharmacology into clinically relevant biomarkers.