Prof. Dr. Frank Noé

Head of Computational Molecular Biology group

FU Berlin
Arnimallee 6
14195 Berlin
+49 (0) 30 838 75354
frank.noe@fu-berlin.de
Website


Research focus

Molecular Dynamics
Cellular dynamics
Reaction-diffusion simulation
Scientific computing
High-performance computing
Time series analysis
Machine learning

Projects as a project leader

  • CH17

    Hybrid reaction-diffusion / Markov-state model of systems with many interacting molecules

    Prof. Dr. Frank Noé / Prof. Dr. Christof Schütte

    Project heads: Prof. Dr. Frank Noé / Prof. Dr. Christof Schütte
    Project members: -
    Duration: -
    Status: running
    Located at: Freie Universität Berlin

    Description

    While simulations of detailed molecular structure, e.g. using atomistic or coarse- grained MD simulation is able to describe the evolution of molecular systems at length/timescales of nanometers/milliseconds, we require a way to bridge from the molecular scale to large-scale/long-time evolutions of molecular superstructures such as actin networks on the scale of micrometers/hours. Such time- and lengthscales while still maintaining some structural, and importantly single-molecule resolution, can be covered by particle-based reaction-diffusion simulations. Molecular kinetic models of small parts of the overall machinery (single molecules and small complexes) can be parametrized with high-throughput MD simulations, enhanced sampling simu- lations, possibly by incorporating constraints from experimental data. In order to ex- plore the long-range and long-time behavior of mixtures and superstructures of many molecules, we set out ot develop a rigorous and computationally efficient coupling be- tween molecular kinetics models and particle-based reaction-diffusion dynamics (Fig. 1).

  • CH19

    Estimating Dynamics of Macromolecular Systems by Low Rank Approximation Techn

    Priv.-Doz. Dr. Konstantin Fackeldey / Prof. Dr. Frank Noé / Prof. Dr. Reinhold Schneider / Hao Wu

    Project heads: Priv.-Doz. Dr. Konstantin Fackeldey / Prof. Dr. Frank Noé / Prof. Dr. Reinhold Schneider / Hao Wu
    Project members: -
    Duration: 01.06.2017 - 31.12.2018
    Status: running
    Located at: Freie Universität Berlin / Technische Universität Berlin

    Description

    The dynamics of a molecular system can be described by the propagation of probabilities. The project aims at estimating coarse grained models of probability densities for molecular dynamics (MD) by nonlinear projections from a high dimensional space onto a low dimensional space. Molecular processes such as protein kinetics from all-atom simulations and the like suffer from the high dimensionality of the underlying space. To overcome this, projections from the high dimensional space onto a low dimensional space have been introduced, such that the system can be described on a coarser scale by using less degrees of freedom. In the present project we apply low rank tensor approximations, to tackle the curse of dimensions. We will use Observable Operator models (OOM) to estimate the dynamics using data from short time simulation.

  • CH-AP7

    Efficient calculation of slow and stationary scales in molecular dynamics

    Prof. Dr. Frank Noé

    Project heads: Prof. Dr. Frank Noé
    Project members: -
    Duration: -
    Status: running
    Located at: Freie Universität Berlin

    Description

    Molecular dynamics (MD) simulation is a technique that may aid in the understanding of fundamental processes in biology and chemistry, and has important technological applications in pharmacy, biotechnology, and nanotechnology. Many complex molecular processes are of multi-scale nature in that they have timescales spanning the range 10−15 s to 1 s, often with no pronounced gap that would permit efficient coarse-grained time integration. Molecular dynamics is a Markov process in a high-dimensional state space. The dom- inant timescales and their associated transitions between metastable (long-lived) states are given by the eigenvalues and eigenfunctions of the transfer operator of the Markov process. These dominant eigenvalues and eigenfunctions therefore need to be approxi- mated. The introduction of Markov state models (MSMs) to molecular simulation in the past few years has been a breakthrough in providing the ability to perform such an approx- imation. An MSM consists of a discretization of the molecular state space into sets, often found by geometric clustering of available simulation data, and a matrix of tran- sition probabilities between them, estimated from the same simulation data. This is an estimation of a set discretization of the transfer operator. Despite their success, MSMs currently suffer from two fundamental problems: 1. Discretization Problem: When the initial discretization for the MSM, based on geometric distances in the data, is poor, the results will be spurious, resulting in numerical unreliability. When the user is interested in approximating a sizable number (e.g. 10 − 100) of slow processes with high accuracy, the common practice to use data-driven geometric clustering methods may not be a viable approach. 2. Sampling Problem: MSMs contain only information of states that have been visited and transitions that have occurred in the simulation data. While the slowest events may occur on timescales of seconds, affordable simulation lengths are on the order of microseconds. Thus, MSM construction suffers from a severe sampling problem. Both problems are coupled, and we now set out to develop a concise mathematical and algorithmic framework to address them.

  • CH-AP10

    Multiscale modeling and simulation for spatiotemporal master equations

    Prof. Dr. Frank Noé / Prof. Dr. Christof Schütte

    Project heads: Prof. Dr. Frank Noé / Prof. Dr. Christof Schütte
    Project members: -
    Duration: 01.10.2014 - 30.06.2018
    Status: running
    Located at: Freie Universität Berlin

    Description

    Accurate modeling of reaction kinetics is important for understanding the functionality of biological cells and the design of chemical reactors. Depending on the particle con­centrations and on the relation between particle mobility and reaction rate constants, different mathematical models are appropriate. In the limit of slow diffusion and small concentrations, both discrete particle numbers and spatial inhomogeneities must be taken into account. The most detailed root model consists of particle-based reaction-diffusion dynamics (PBRD), where all individual par­ticles are explicitly resolved in time and space, and particle positions are propagated by some equation of motion, and reaction events may occur only when reactive species are adjacent.
    For rapid diffusion or large concentrations, the model may be coarse-grained in dif­ferent ways. Rapid diffusion leads to mixing and implies that spatial resolution is not needed below a certain lengthscale. This permits the system to be modeled via a spa­tiotemporal chemical Master equation (STCME), i.e. a coupled set of chemical Master equations acting on spatial subvolumes. The STCME becomes a chemical Master equa­tion (CME) when diffusion is so fast that the entire system is well-mixed. When particle concentrations are large, populations may be described by concentrations rather than by discrete numbers, leading to a PDE or ODE formulation.

    Many biological processes call for detailed models (PBRD, ST-CME or CME), but these models are extremely costly to solve. Ef.cient mathematical and computational methods are needed in order to approximate the solutions of these models with some guaranteed accuracy level. An approach to optimal or ef.cient switching between different models is, as yet, missing.
    In this project, we will set out to develop a multiscale theory for reaction kinetics processes, starting from a consistent and well-de.ned formulation of PBRD models, and including spatial scaling (PBRD <-> ST-CME <-> CME) coupled to population scaling (CME <-> ODE). In particular, we aim at providing solutions for the problematic cases of having particles at diverse copy numbers (CME . ODE) and at least some slowly diffusing particles (PBRD <-> CME <-> STCME). The cascades of scales in these scenarios and efficient approximation strategies will be explored.

    http://sfb1114.imp.fu-berlin.de/research/index.php?option=com_projectlog&view=project&id=12
  • CH-AP12

    pcCells - physicochemical principles of cellular information processing

    Prof. Dr. Frank Noé

    Project heads: Prof. Dr. Frank Noé
    Project members: -
    Duration: 01.01.2013 - 31.12.2017
    Status: running
    Located at: Freie Universität Berlin

    Description

    Biological cells are able to perform complex signal transduction tasks quickly, energy-efficiently and yet in a robust and noise-tolerant manner. These signal transduction tasks rely on intracellular information processing mechanisms in which chemical signals are sent, transmitted and received and the state of the overall machinery is stored in chemical or conformational switches. The physicochemical principles of information processing in cells is still not well understood, owing to fundamental restrictions in resolution in experiments and in sampling of molecular dynamics simulations. Here, we will develop new simulation methods based on adaptive molecular dynamics and Markov models. These methods, together with new statistical mechanical theories and single-molecule experimental analyses will be employed to investigate the molecular basis of intracellular signal processing mechanisms. Central to our proposal is the hypothesis that intracellular signal processing relies on spatiotemporal order of molecules arising from dynamical sorting. This hypothesis will be tested using examples of protein-ligand and protein-protein sorting in neuronal signalling. The proposed project is highly multidisciplinary, involving physical chemistry, computer science, mathematics and biology.

    http://compmolbio.biocomputing-berlin.de/index.php/projects/80-erc
  • CH-AP14

    Conformational dynamics of biomolecules: Reconciling simulation and experimental data

    Prof. Dr. Frank Noé

    Project heads: Prof. Dr. Frank Noé
    Project members: -
    Duration: 01.01.2012 - 31.12.2018
    Status: running
    Located at: Freie Universität Berlin

    Description

    We develop methods for constructing kinetic models of biomolecular conformation dynamics from single-molecule experimental data, or by reconciling kinetic experimental data and molecular dynamics simulation data. In the present funding period we aim at developing an approach to directly compute MSMs from single-molecule experiments, with the following main objectives:
    1. Estimate conformation dynamics (eigenvalues, eigenvectors of the underlying Markovian dynamics) directly from single-molecule trajectories.
    2. Quantify the estimation errors of 1.
    3. Applications to optical tweezer data (collaboration with Susan Marqusee).
    4. Use 1. to compute improved Markov state models; Applications to simulation data.
    5. Provide publicly available software implementation.


    http://compmolbio.biocomputing-berlin.de/index.php/projects/91-dfg825-2-2
  • CH-AP13

    Adaptive Konformationsdynamik mit Anwendung auf Rhodopsinaktivierung

    Prof. Dr. Frank Noé

    Project heads: Prof. Dr. Frank Noé
    Project members: -
    Duration: 01.07.2012 - 30.06.2015
    Status: completed
    Located at: Freie Universität Berlin

    Description

    Rare molecular events such as folding of proteins or nucleic acids, ligand binding, conformational changes or macromolecular aggregation are the basis of all life processes. Besides experimental techniques, molecular dynamics (MD) simulation is an established tool to analyze such processes. However, the usefulness of MD for investigating biological processes is limited by the sampling problem: Due to the high computational effort involved in simulating biomolecules at atomistic resolution, the accessible simulation times are much too short to find the biologically relevant conformations and make statistically reliable statements about transition rates. This problem also hinders the improvement of molecular models towards the reliable prediction of experimental observables. In the proposed work we will develop an adaptive conformation dynamics (ACD) which facilitates the simulation of slow biomolecular processes on small CPU clusters using atomistic models. This method will be applied in order to elucidate the detailed structural mechanism of the activation of the G-protein coupled receptor Rhodopsin.

    http://compmolbio.biocomputing-berlin.de/index.php/projects/93-dfg825-3-1