Search papers, labs, and topics across Lattice.
The study introduces UBio-MolFM, a foundation model capable of simulating biomolecular dynamics with density functional theory (DFT) accuracy for systems containing up to 100,000 atoms. By leveraging a model trained on 160 million quantum-chemical labels, it achieves near-linear computational costs while maintaining a force error of approximately 20 meV/{\AA}. Key findings include the accurate reproduction of water's X-ray structure and ion hydration, as well as the successful simulation of a 108,964-atom KcsA channel, demonstrating the model's potential to overcome the limitations of traditional first-principles simulations.
First-principles simulations can now accurately model biomolecular systems with over 100,000 atoms at a fraction of the cost.
Ion conduction, membrane permeation and metal recognition hinge on electronic structure, yet first-principles simulation reaches only hundreds of atoms. UBio-MolFM lifts that ceiling: a foundation model trained on 160 million quantum-chemical labels, its receptive field spanning non-covalent distances at near-linear cost. The barrier is cost, not principle. One untuned potential keeps force error near 20 meV/{\AA} past a thousand atoms, reproduces water's X-ray structure and ion hydration, and holds an RNA Mg$^{2+}$ site without ion-specific parameters. Cyclosporine A pays 3.5 kcal/mol in water for its permeable conformer, gated by one kinetically asymmetric hydrogen bond that a fixed-charge model flattens. In a 108,964-atom KcsA channel on one GPU, the relaxed four-ion column is anhydrous in all five replicas, in direct contact in four---the knock-on geometry ten fixed-charge simulations never form. It remains orders of magnitude costlier. Where electronic structure decides the answer, first-principles simulation is in reach.
