Title:	Scaling Transferable Implicit Transfer Operators
DNr:	Berzelius-2026-213
Project Type:	LiU Berzelius
Principal Investigator:	Simon Olsson <simonols@chalmers.se>
Affiliation:	Chalmers tekniska högskola
Duration:	2026-07-09 – 2027-02-01
Classification:	10210
Homepage:	https://psolsson.github.io/
Keywords:

Abstract

Computing properties of molecular systems rely on estimating expectations of the (unnormalized) Boltzmann distribution. Molecular dynamics (MD) is a broadly adopted technique to approximate such quantities. However, stable simulations rely on tiny integration time steps (10^(−15) s), whereas convergence of some moments, e.g., binding free energy or rates, might rely on sampling processes on time scales as long as 10^(−1) s, and these simulations must be repeated for every molecular system independently. We recently proposed Implicit Transfer Operator (ITO) Learning [1], a framework to learn surrogates of the simulation process with multiple time resolutions. Our initial work shows that ITO models can simulate challenging molecular systems, such as fast-folding proteins, with time steps at least six orders of magnitude larger than traditional molecular dynamics [1]. We have show that these models can be trained to generalize to small molecules and peptides [2] as well as small coarse-grained single domain proteins [3] in previous allocations, and we have built the first steps towards an all-atom single-domain protein model (unpublished). In this allocation, we plan to scale this to arbitrarily sized single-chain proteins with all-atom resolution using latent generative models coupled with a low-rank attention mechanism. Since we are now reaching a stage where comparison to the ground truth simulation model is intractable, since numerical simulations are costly, we will do a `sim2real' transfer and use our approach [4] to calibrate this new model to all available biophysical data. [1] Mathias Schreiner, Ole Winther and Simon Olsson. Implicit Transfer Operator Learning: Multiple Time-Resolution Surrogates for Molecular Dynamics. 37th Conference on Neural Information Processing Systems (NeurIPS 2023). https://arxiv.org/abs/2305.18046 [2] Viguera Diez J, Schreiner M, Olsson S . "Transferable Generative Models Bridge Femtosecond to Nanosecond Time-Step Molecular Dynamics" Science Advances. [3] Antoniadis P, Pavesi B, Olsson S*, Winther O* "Protein Language Model Embeddings Improve Generalization of Implicit Transfer Operators" ICML 2026 [4] Kolloff C, H\"oppe T, Angelis E, Schreiner M, Bauer S, Dittadi A, Olsson S* "Minimum Excess Work Guidance: Score-Based Sampling with Experimental Data or Sparse Restraints" J. Chem. Theory. Comput. (2026)