Deep Generative Markov State Models

By Hao Wu, Andreas Mardt, Luca Pasquali, Frank No'e. 2018

68 citations recorded by the source corpus.

We propose a deep generative Markov State Model (DeepGenMSM) learning framework for inference of metastable dynamical systems and prediction of trajectories. After unsupervised training on time series data, the model contains (i) a probabilistic encoder that maps from high-dimensional configuration space to a small-sized vector indicating the membership to metastable (long-lived) states, (ii) a Markov chain that governs the transitions between metastable states and facilitates analysis of the long-time dynamics, and (iii) a generative part that samples the conditional distribution of configurations in the next time step. The model can be operated in a recursive fashion to generate trajectories to predict the system evolution from a defined starting state and propose new configurations. The DeepGenMSM is demonstrated to provide accurate estimates of the long-time kinetics and generate valid distributions for molecular dynamics (MD) benchmark systems. Remarkably, we show that DeepGenMSMs are able to make long time-steps in molecular configuration space and generate physically realistic structures in regions that were not seen in training data.

Subfields: Materials Chemistry, Molecular Biology, Statistical and Nonlinear Physics, Generative Models, Probabilistic & Bayesian Methods

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Materials Chemistry

1. Frank No'e
2. K. Müller
3. Steven L. Brunton
4. Cecilia Clementi
5. Alexandre Tkatchenko
6. Petros Koumoutsakos
7. Bernd R. Noack
8. Kristof T. Schütt

In Molecular Biology

1. Frank No'e
2. Cecilia Clementi
3. K. Müller
4. Alexandre Tkatchenko
5. Steven L. Brunton
6. Petros Koumoutsakos
7. Bernd R. Noack
8. Hao Wu

In Statistical and Nonlinear Physics

1. Steven L. Brunton
2. J. Nathan Kutz
3. Bernd R. Noack
4. Cecilia Clementi
5. Petros Koumoutsakos
6. Frank No'e
7. K. Müller
8. Alexandre Tkatchenko

In Generative Models

1. Frank No'e
2. Steven L. Brunton
3. K. Müller
4. Alexandre Tkatchenko
5. Cecilia Clementi

In Probabilistic & Bayesian Methods

1. Steven L. Brunton
2. K. Müller
3. Alexandre Tkatchenko
4. Frank No'e
5. Cecilia Clementi
6. Bernd R. Noack
7. Mattia Segu
8. Petros Koumoutsakos

Citations in the corpus

• A General Framework for Uncertainty Estimation in Deep Learning (2019, background): “Similarly, probabilistic deep state space models retrieve data uncertainty in sequential data, and use it for learning-based filtering [11], [12].”
• Machine learning for molecular simulation (2019, methodology): “Analysis and simulation of MD trajectories has also been affected by ML, for instance for the definition of optimal reaction coordinates [19, 20, 21, 22, 23, 24], the estimate of free energy surfaces [25, 26, 27, 22], the construction of Markov State Models [21, 23, 28], and for enhancing MD samplin”
• Machine Learning for Fluid Mechanics (2019, methodology): “Generative adversarial networks (GANs) (Goodfellow et al. 2014) are also being used to infer dynamical systems from data (Wu et al. 2018).”