These are preliminary reports that have not been peer-reviewed. They should not be regarded as conclusive, guide clinical practice/health-related behavior, or be reported in news media as established information. For more information, please see our FAQs.
2 files

Targeted Adversarial Learning Optimized Sampling

revised on 22.05.2019, 18:47 and posted on 22.05.2019, 21:06 by Jun Zhang, Yi Isaac Yang, Frank Noé
Abstract Boosting transitions of rare events is critical to modern-day simulations of complex dynamic systems. We present a novel approach to modify the potential energy surface in order to drive the system to a user-defined target distribution where the free energy barrier is lowered. The new approach, called targeted adversarial learning optimized sampling (TALOS) combines the strengths of statistical mechanics and deep learning. By casting the enhanced sampling problem as a competing game between a real sampling engine and a virtual discriminator, TALOS enables unsupervised construction of bias potential on an arbitrary dimensional space and seeks for an optimal transport plan that transforms the system into target. Through multiple experiments we show that on-the-fly training of TALOS benefits from the state-of-art optimization techniques in deep learning, thus is efficient, robust and interpretable. Additionally, TALOS is closely connected to actor-critic reinforcement learning, giving rise to a new approach to manipulating the Hamiltonian systems via deep learning.


Alexander von Humboldt Fellowship

European Commission (ERC CoG 772230 “ScaleCell” and ERC-2014-ADG-670227 “VARMET)

Swiss National Science Foundation (NCCR MARVEL 51NF40_141828)


Email Address of Submitting Author


Freie Universität Berlin



ORCID For Submitting Author


Declaration of Conflict of Interest

The authors declare no conflict of interest.

Version Notes

Version 3.0


Read the published paper

in The Journal of Physical Chemistry Letters