Scaffold Fusion and SAR Transfer with a Chemical Language Model Generates Novel Liver X Receptor Modulators.

Liver X receptors (LXRs) are promising targets for metabolic disorders including atherosclerosis and metabolic dysfunction-associated steatotic liver disease (MASLD). In this study, we employed a chemical language model (CLM) for LXR modulator design in an explorative fashion and observed that structural features from different LXR modulator templates were merged, and structure-activity relationship (SAR) knowledge was transferred. The generated computational designs demonstrated LXR modulation with diverse activity profiles and selective modulator properties, including a promising lipolytic activity of an inverse LXR agonist in an in vitro MASLD model that warrants its further development to improve ADME properties.