Insilico Medicine has nominated ISM1354, an AI-designed small molecule aimed at the glucose-dependent insulinotropic polypeptide receptor, as a preclinical candidate for obesity and Type 2 diabetes.
The compound came out of Chemistry42, Insilico’s generative chemistry platform. The company leaned on protein-ligand structures, an early liver-injury prediction filter and free energy perturbation modeling across repeated design cycles.
Preclinical work showed oral bioavailability of 75% to 104% across mice, rats, dogs and monkeys. At the same dose, the compound produced at least 18 times greater plasma exposure than a clinical-stage benchmark, Insilico said.
Safety data also looked differentiated. The company reported weaker OATP1B1 inhibition than the benchmark and a wider hepatocyte safety margin, with little effect on cell viability up to 200 micromolar. Non-GLP monkey toxicology showed roughly a 45-fold safety margin.
Why pair a GIPR drug with the blockbuster GLP-1 class? The hope is that combining them strips fat without eroding lean muscle, a priority for societies facing more frailty in old age.
