Skip to Main Content
From Computational Biophysics to Systems Biology (CBSB12)
June 3-5, 2012  Knoxville, Tennessee
  home  
ORNL link
OU link
UT link
ORNL link

Sponsored by:
ORNL link
Invited Speakers:

Keynote Speakers:

Invited Speakers:

Abstracts:
  • Exploring Landscapes for Protein Folding and Binding Using Replica Exchange Dynamics, Kinetic Networks and Markov State Models
    Ronald M. Levy

    Department of Chemistry and Chemical Biology, and BioMaPS Institute for Quantitative Biology, Rutgers University

    Advances in computational biophysics depend critically on the development of accurate effective potentials and powerful sampling methods to traverse the rugged energy landscapes that govern protein folding, binding and fitness. I will review work in my lab over the last few years concerning the construction of all-atom effective potentials for proteins and multi-scale methods for simulating their folding and binding on long time scales. Replica exchange (RE) is a generalized ensemble molecular simulation method for accelerating the exploration of free-energy landscapes which define many challenging problems in computational biophysics, including protein folding and binding. We have clarified some of the obstacles to obtaining converged thermodynamic information from RE simulations. I will discuss these issues and new multi-scale approaches to recover protein folding rates and pathways for folding and binding using the combined power of replica exchange, kinetic network models with flux analysis, and effective stochastic dynamics.
    Back to the top of the page

  • Theory and Simulation of Biomolecular Systems: Surmounting the Challenge of Bridging the Scales
    Gregory A. Voth

    Department of Chemistry, James Franck Institute, Institute for Biophysical Dynamics, and Computation Institute, University of Chicago

    A multiscale theoretical and computational methodology will be presented for studying biomolecular systems across multiple length and time scales. The approach provides a systematic connection between all-atom molecular dynamics, coarse-grained modeling, and mesoscopic phenomena. At the heart of the approach is a method for deriving coarse-grained models from protein structures and their underlying molecular-scale interactions. This particular aspect of the work has strong connections to the theory of renormalization, but it is more broadly developed and implemented for heterogeneous systems. A critical component of the methodology is also its connection to experimental structural data such as cryo-EM or x-ray, thus making it “hybrid” in its character. Applications this overall multiscale approach to study key features of large multi-protein complexes such the HIV-1 virus capsid, the entire HIV-1 immature virion, actin filaments, and protein-mediated membrane remodeling will be presented as time allows.
    Back to the top of the page

  • Simulating the transport of a cellulose chain through the cellulase catalytic tunnel
    Xiaolin Cheng

    Center for Molecular Biophysics, Okaridge National Laboratory



    The degradation of cellulosic biomass to sugars both in nature and in biorefineries is primarily accomplished by enzymes such as the Family 7 cellobiohydrolase (Cel7A) from Trichoderma reesei. Therefore, a molecular-level understanding of the mechanisms of cellulose-degrading enzymes is critical to developing improved technologies for biofuel production. We have run extensive molecular dynamics simulations of a single cellulose chain inside the Cel7A catalytic tunnel to understand the structural basis for processivity. In particular, we have investigated the initial binding of a cellulose chain into the catalytic tunnel of Cel7A. In multiple unbiased simulations, the cellulose chain spontaneously diffuses into the tunnel by the length of a cellobiose unit. Further free energy calculations reveal an overall downhill free energy profile for the initial threading, which becomes progressively flat and rugged toward the active site, highlighting a strong interplay between processivity and catalysis. The free energy profiles for different cellulose chain orientations show a clear thermodynamic preference for the reducing end binding at the -5 subsite, suggesting that the selective hydrolysis of cellulose from the reducing end might be partially achieved through the preferential initial binding of cellulose to Cel7A.
    Back to the top of the page

  • A Computational Platform for Protein Structure Determination Integrating Limited Experimental Data
    Jens Meiler

    Computational Chemical and Structural Biology, Vanderbilt University

    I will present BCL::Fold, a new algorithm for de novo prediction of complex and large protein topologies by assembly of secondary structure elements. The method was designed to integrate experimental data from NMR, EPR, EM, SAXS experiments, or combinations thereof. These data sets often provide more readily restraints for regions of defined secondary structure. I will present examples for atomic-detail structure elucidation from medium resolution cryo-EM density maps (Figure 1) and paramagnetic restraints from NMR spectroscopy. Briefly: Computational de novo protein structure prediction is limited to small proteins of simple topology. The present work explores an approach to extend beyond the current limitations through assembling protein topologies from idealized α-helices and β-strands. The algorithm performs a Monte Carlo Metropolis simulated annealing folding simulation. It optimizes a knowledge-based potential that analyzes radius of gyration, β-strand pairing, secondary structure element packing, amino acid pair potential, amino acid environment, and loop closure. Discontinuation of the protein chain favors sampling of non-local contacts and thereby creation of complex protein topologies. The folding simulation is accelerated through exclusion of flexible loop regions further reducing the size of the conformational search space. The algorithm is benchmarked on 66 proteins with lengths between 83 and 293 amino acids. For 61 out of these proteins the best SSE-only models obtained have an RMSD100 below 8.0 Å and recover more than 20% of the native contacts. The algorithm assembles protein topologies with up to 215 residues and a relative contact order of 0.46. BCL::Fold includes a modified scoring function for the assembly of membrane proteins. BCL::Fold is typically combined with Rosetta refinement algorithms to arrive at proteins models accurate at atomic detail.
    Back to the top of the page

  • Knowledge-Based Structural Approaches for Predicting Hot Spots of Protein Binding and Allostery
    Julie Mitchell

    Departments of Mathematics and Biochemistry, University of Wisconsin - Madison

    Using information derived from protein structures, it is possible to predict amino acid positions where mutations will have a deleterious effect on protein binding or allosteric communication. The KFC2 model captures 80% of alanine scanning mutagenesis hot spots, which result in a binding energy increase of at least 2 kcal/mol. A unique feature of the model is a local plasticity feature that suggests whether a change in sequence can be accommodated through local sidechain rearrangements. A different plasticity measure, known as local structural entropy, is a dominant feature in our AlloSIND model for allosteric hot spots that lie between the effector and active sites of allosteric proteins. One possible interpretation is that rigidity of internal protein secondary structure prevents an allosteric protein from absorbing the impact of effector binding locally, resulting in longer range conformation effects.
    Back to the top of the page

  • Free energy landscapes of polyproline and polyglutamine peptides
    Celeste Sagui

    Department of Physics, North Carolina State University

    We use enhanced sampling techniques (the Adaptively Biased Molecular Dynamics (ABMD) method, multiple walkers, replica exchange, steered MD, and various combinations thereof) to study peptide systems whose conformational space cannot be sampled with regular MD simulations. These include transitions between the PPII and PPI forms of polyproline (polyP); proline-rich guest/host peptides; polyglutamine (polyQ), and polyQ-polyP systems. Several statistical techniques allow us to explore the atomic mechanisms that underlie various experimental observations: the apparent PPII propensity of guest amino acids in polyP-rich peptides; the properties that may favor aggregation in polyQ systems; and the suppression of aggregation of polyQ by the addition of a C-terminal polyP peptide.
    Back to the top of the page

  • Role of hydration and confinement in protein folding and aggregation
    John E. Straub

    Department of Chemistry, Boston University

    Reverse micelles provide an environment in which the number of water molecules and overall cavity size may be "tuned," by adjusting the water-to-surfactant ratio, allowing, in principle, the role of hydration and confinement on protein folding and aggregation to be systematically studied. We have used molecular dynamics simulations to explore the structure and dynamics of the alanine-rich AKA2 peptide and aggregation-prone fragments of amyloid proteins in reverse micelle environments. The dependence of the peptide-micelle interaction on capping of the N- and C-termini and the nature of the force field is explored. The time scales for fluctuations in the reverse micelle shape and surface area are characterized and compared with the results of more idealized spherical micelle models. The results suggest that an understanding of the detailed nature of protein-surfactant interactions can be essential to the interpretation of studies of protein folding and aggregation in reverse micelle environments.
    Back to the top of the page

  • From Computational Docking to Exploration of Biochemical Pathways
    Jerome Baudry

    Department of Biochemistry, Cellular and Molecular Biology, University of Tennessee, Knoxville

    Virtual (in-silico) docking of small molecules in protein targets is a popular and successful approach to discover molecules capable of binding in proteins. We describe how docking is used to investigate biochemical pathways, focusing on how P450s can turn environmental molecules into estrogenic pollutants by increasing their binding to the estrogen receptor alpha target. We also describe ongoing developmental work to use supercomputing architectures efficiently to perform massive docking of very large chemical databases against a large number of protein targets.
    Back to the top of the page

  • Modeling disordered states of Proteins: Are structural models of the unfolded state correct?
    Collin Stultz

    EECS & HST, Massachusetts Institute of Technology

    The characterization of intrinsically disordered proteins is challenging because accurate models of these systems require a description of both their thermally accessible conformers and the associated relative stabilities or population weights. These structures and weights are typically chosen such that calculated ensemble averages agree with some set of pre-specified experimental measurements; however, the large number of degrees of freedom in these systems typically leads to multiple conformational ensembles that are degenerate with respect to any given set of experimental observables. Moreover, our recent work demonstrates that estimates of the relative stabilities of conformers within an ensemble are often incorrect when one does not account for the underlying uncertainty in the estimates themselves. Therefore, we have developed a method for modeling the conformational properties of disordered proteins that estimates the uncertainty in the weights of each conformer. A unique and powerful feature of the approach is that it provides a built-in error measure that allows one to assess the accuracy of the ensemble. Using this approach we constructed an ensemble that characterizes the accessible states of the IDPs, tau protein, alpha-synuclein and abeta; i.e., proteins that play a role in several neurodegenerative disorders. These data led to new insights into intramolecular interactions that may play a role in promoting IDP self-association - a process which has been linked to neuronal death and dysfunction in patients with Alzheimer's disease. We further demonstrate that these data may be used in the initial stages of a strategy to design ligands that prevent IDP aggregation. More generally, we derive an order parameter that quantifies the extent of disorder within a protein. Although protein disorder is normally thought of as a binary phenomenon (i.e., a protein is either disordered or not), we suggest that the concept of protein disorder should be treated like a continuous variable, and that not all unfolded states are created equal.
    Back to the top of the page

  • Recent development and application of continuous constant pH molecular dynamics
    Jana Shen

    Chemistry, University of Oklahoma

    Development of the constant pH molecular dynamics techniques has opened a door to atomically detailed studies of dynamic processes coupled to protonation/deprotonation. Here we discuss the most recent development of the continuous constant pH molecular dynamics (CpHMD) technique and application studies for gaining novel insights into ionization-coupled conformational phenomena in biology and chemistry. We show that CpHMD simulations offer, for the first time, thermodynamic description of coupled protonation and conformational equilibria for proteins. We will also discuss other applications such as pH titration of micelles and pH-dependent phase transitions.
    Back to the top of the page

  • Enhanced sampling simulations of DNA and protein-DNA complexes
    Arjan van der Vaart

    Department of Chemistry, University of South Florida

    The flexibility of long DNA is well-described by the worm-like chain model, but at the small-length scales this model breaks down. Moreover, several experiments suggest that short DNA has an increased flexibility. We performed enhanced sampling simulations of short DNA strands to assess its flexibility. Our results indicate that the stiffness of DNA decreases upon strong bending. We will also discuss enhanced sampling simulations to show how proteins and ligands modulate the DNA structure.
    Back to the top of the page

  • High throughput identification and druggability analysis of protein binding sites
    Sandor Vajda

    Department of Biomedical Engineering and Chemistry, Boston University

    Our lab has developed the computational solvent mapping method of determining binding hot spots of proteins (1). The method globally samples the surface of a target protein using small organic molecules as probes, finds favorable positions, clusters the conformations, and ranks the clusters on the basis of the average energy. The regions that bind several probe clusters predict the binding hot spots, in good agreement with experimental results. Solvent mapping can be used to solve two important problems. First, it achieves higher accuracy than any other method in the identification of ligand binding sites on unbound protein structures (2). Second, the mapping results provide information for assessing druggability, i.e., the ability of a protein to bind drug sized ligands with high affinity (3). While both applications were very successful, they required lengthy calculations and hence were originally demonstrated on small benchmark sets. We describe our work toward modifying the methods such that applications to large sets of proteins would become computationally feasible, enabling some general conclusions on the binding properties of proteins.
    Back to the top of the page