ReviewConformational sampling for the impatient
Introduction
An intellect which at any given moment knew all the forces that animate nature and the mutual positions of the being that comprise it, if this intellect were vast enough to submit its data to analysis, could condense into a single formula the movement of the greatest bodies of the universe and that of the lightest atom: for such an intellect nothing could be uncertain; and the future just like the past would be present before our eyes [1], [2].
After nearly two centuries, this Newtonian–Laplacian statement seems over-optimistic, if valid at all. The collective intellect of the scientific community, even when given the crystallographic positions of atoms in a protein, finds itself scrambling for barely enough computing power to predict the behaviour thereof.
The motivation for this review is to list the new computational methods for conformational sampling, given structural knowledge of only one conformational state of a biomolecule. The community now faces new challenges, such as the need to simulate larger biopolymers and even supramolecular assemblages [3], [4], but conventional molecular dynamics (MD) method is showing its limitations [5], [6]. Though new methods have indeed sprung up to take on the challenges in their stride, the last reviews covering similar grounds are now a few years old [7], [8], [9], [10].
The title of this review may be taken in two ways: First, the impatient researcher, interested in the dynamics and function of a protein [11], may wish to use a new, faster method than conventional MD to sample the conformations of a biomolecular system so as to save a few weeks of computing time; second, the reviewer writes with the hope that he may save the researcher a few days of plowing through in the literature looking for the appropriate sampling method.
Before we start, readers new to the metaphor of energy landscapes of biomolecules are encouraged to familiarize themselves using the papers by Frauenfelder et al. [12], [13], [14]. Principal component analysis (PCA), also called essential dynamics [15], [16], [17], has become a standard analysis tool for MD trajectories. It has been used to demonstrate the capability [18] and expose the inadequacies [19], [20] of MD in conformational sampling. Some of the most powerful energy landscape visualization [21] and enhanced sampling (reviewed below) methods are built upon the foundation of PCA. Therefore, a good understanding of PCA is essential. In these matters, several chapters of the book edited by Becker et al. [22] are helpful.
As much as possible, this review aims to include recent application of methods to specific biomolecular systems, with a bent towards the functional dynamics of proteins. Whenever something has not been—but can be—tried, hints are placed to inspire the reader. Methods for determining the reaction path given the initial and final conformations are outside our scope; the reader is directed to two recent reviews [23], [24].
Section snippets
Multiple timestep methods
The development of multiple timestep MD has been going on for a few years [25], the most popular being the RESPA (reference system propagation algorithm) methods [26], [27], [28]. Applications to biological systems have been tested, showing that carefully using an appropriate method with an outer timestep of approximately 5 fs can be safe—that is, giving a stable and useful trajectory–whilst affording a 2–3-fold speed-up [29]. Notably, the tests have included some lipid membrane systems [30],
Atomic-scale biased methods: fighting it out of the minimum
Unlike methods in the previous section, the ones following are biased in a struggle to get out of the energy minimum and achieve sampling a wider space in the conformational space. Much development has occurred in the field of material science [54], [55], [56], [57]; most of the methods covered here, however, have been designed with protein dynamics and functional studies in mind, to search for interesting spots (minima or conformational transitions) in the energy landscape of biopolymers.
CONCOORD: a non-dynamical method of generating conformation sets
The popular CONCOORD (‘from constraints to co-ordinates’) method [72] has now been applied to a wide variety of proteins, some as large as the GroEL chaperonin [73]. Unlike MD or most of the other methods reviewed here, it is not a dynamical method—the set of conformation generated cannot be considered as a timeseries. Rather, the resulting set contains conformations satisfying a list of distance criteria. It is also notable that the idea of energy does not take a direct role; this perhaps
Network models
The Gaussian network model [74], [75] has a residue-level (as opposed to atomic-level) model of a protein. The residues are connected by harmonic potentials governed by a single Hookean parameter (anharmonic terms are added in a later variety [76]).
It very aptly reproduces isotropic temperature (B) factors observed in crystallography, and has been applied to proteins such as tryptophan synthase [77], HIV-1 reverse transcriptase [78] and influenza virus hemagglutinin [79]. Clearly, global
Summary
Distinct exhortations are directed to two constituent groups in the scientific community. To those who are eager to sample the conformations of larger proteins and interested in dynamical behaviours, this review commends the RESPA, replica-exchange MD, CONCOORD and Gaussian network methods. For those who are keen to develop better methodologies for conformational sampling, endeavours in the directions of the Langevin, swarming, self-guided MD and MBO(N)D methods may prove effective. May this
Acknowledgements
The reviewer thanks Profs. Mark S.P. Sansom and J. Andrew McCammon, their groups, and the Biotechnology and Biological Sciences Research Council (United Kingdom) for support and the Revd Dr Ian James for bringing the Laplace quote [1], [2] to his attention. He also thanks Mr Oliver Beckstein and Prof. Stephen David Bond for reading the manuscript.
References (86)
- et al.
Enlarging the landscape
Curr. Opin. Struct. Biol.
(2002) Modeling supramolecular assemblages
Curr. Opin. Struct. Biol.
(2002)Long timescale simulations
Curr. Opin. Struct. Biol.
(2000)- et al.
Molecular dynamics simulations
Curr. Opin. Struct. Biol.
(2002) Computational molecular biophysics today: a confluence of methodological advances and complex biomolecular applications
J. Comput. Phys.
(1999)- et al.
Algorithmic challenges in computational molecular biophysics
J. Comput. Phys.
(1999) - et al.
Collective protein dynamics in relation to function
Curr. Opin. Struct. Biol.
(2000) - et al.
Essential dynamics of reversible peptide folding: memory-free conformational dynamics governed by internal hydrogen bonds
J. Mol. Biol.
(2001) - et al.
Analysis of a 10-ns molecular dynamics simulation of mouse acetylcholinesterase
Biophys. J.
(2001) Time-trimming tricks for dynamic simulations: splitting force updates to reduce computational work
Structure
(2001)
Molecular dynamics simulation of a synthetic ion channel
Biophys. J.
Protein dynamics simulations from nanoseconds to microseconds
Curr. Opin. Struct. Biol.
Hierarchy of structure loss in MD simulations of src SH3 domain unfolding
J. Mol. Biol.
Multicopy molecular dynamics simulations suggest how to reconcile crystallographic and product formationdata for camphor enantiomers bound to cytochrome p-450 cam
J. Inorg. Biochem.
New Monte Carlo algorithms for protein folding
Curr. Opin. Struct. Biol.
Protein folding simulations and structure predictions
Comput. Phys. Commun.
Replica-exchange molecular dynamics method for protein folding
Chem. Phys. Lett.
Parallel tempering algorithm for conformational studies of biological molecules
Chem. Phys. Lett.
Numerical calculations using the hyper-molecular dynamics simulation method
Comput. Phys. Commun.
Puddle-jumping: a exible sampling algorithm for rare event systems
Chem. Phys.
Self-guided molecular dynamics in the isothermal–isobaric ensemble
Chem. Phys. Lett.
Conformational changes in the chaperonin GroEL: new insights into the allosteric mechanism
J. Mol. Biol.
Direct evaluation of thermal fluctuations in proteins using a single-parameter harmonic potential
Fold. Des.
Anisotropy of fluctuation dynamics of proteins with an elastic network model
Biophys. J.
Collective motions in HIV-1 reverse transcriptase: examination of exibility and enzyme function
J. Mol. Biol.
Functional motions of inuenza virus hemagglutinin: a structure-based analytical approach
Biophys. J.
Conformational exibility of pyruvate dehydrogenase complexes: a computational analysis by quantized elastic deformational model
J. Mol. Biol.
Molecular dynamics simulations of peptides and proteins with amplified collective motions
Biophys. J.
A Philosophical Essay on Probabilities
Biomolecular dynamics at long timesteps: bridging the timescale gap between simulation and experimentation
Annu. Rev. Biophys. Biomol. Struct.
The role of dynamics in enzyme activity
Annu. Rev. Biophys. Biomol. Struct.
The energy landscape in non-biological and biological molecules
Nature Struct. Biol.
Energy landscape and fluctuations in proteins
Ann. Phys. (Leipzig)
The role of structure, energy landscape, dynamics and allostery in the enzymatic function of myoglobin
Proc. Natl. Acad. Sci. USA
Large-amplitude non-linear motions in proteins
Phys. Rev. Lett.
Essential dynamics of proteins
Proteins
An efficient method for sampling the essential subspace of proteins
J. Biomol. Struct. Dyn.
Principal component analysis and long time protein dynamics
J. Phys. Chem.
Principal coordinate maps of molecular potential energy surfaces
J. Comput. Chem.
Transition path sampling: throwing ropes over rough mountain passes, in the dark
Annu. Rev. Phys. Chem.
Extending the time scale in atomistic simulation of materials
Annu. Rev. Mater. Res.
Cited by (95)
Coarse-grained molecular dynamics simulations study of the conformational properties of single polyelectrolyte diblock copolymers
2020, Biophysical ChemistryCitation Excerpt :Experimental studies show that they can alter the capture rate and can increase the residence time of such polypeptides wihtin the pore which may be required for DNA sequencing. Computer simulation [15] serves as an important tool to study such conformational behaviors of the block polyelectrolytes that change with variation in the counterion valency, the fraction of charged residues, dielectrics, and salt concentration. In this paper, we study the conformational properties of single di block polyelectrolyte using coarse-grained molecular dynamics (MD) simulations.
Molecular dynamics and simulation
2018, Encyclopedia of Bioinformatics and Computational Biology: ABC of BioinformaticsThe impact of molecular dynamics on drug design: Applications for the characterization of ligand-macromolecule complexes
2015, Drug Discovery TodayCitation Excerpt :Therefore, efficient searching and sampling of the conformational space can be extremely difficult, limiting the accuracy of MD simulations. Several methods have been proposed to enhance the sampling [29,102]. They can be divided into four classes: (i) modification of the potential energy surface; (ii) multicopy searching and sampling; (iii) scaling of system parameters; and (iv) reduction of the number of degrees of freedom.
Computational study on the conformations of CD38 and inclusion complexes of some lower-size large-ring cyclodextrins
2014, Journal of Molecular StructureCitation Excerpt :Of particular importance when examining the modes of conformational interconversions and the dynamics of the intramolecular self-organization is the monitoring of the appearance and the disappearance of band-flips during longer simulation intervals (variation with the simulation time of the flip angle, O3(n)⋯C4(n)⋯C1(n + 1) ···O2(n + 1) – the dihedral flip between secondary hydroxyls of adjacent glucoses). Principal component analysis (PCA), also called quasiharmonic analysis or essential dynamics method [55–57], is a very useful analysis tool for MD trajectories [58–67]. This statistical method allows trajectories obtained from MD simulations to be analyzed by reducing the degrees of freedom of the system to lower dimensions [61].
Advanced replica-exchange sampling to study the flexibility and plasticity of peptides and proteins
2013, Biochimica et Biophysica Acta - Proteins and ProteomicsCitation Excerpt :Conformational transitions between stable states can therefore be rare events even on the time scale of hundreds of nanoseconds to microseconds that can currently be covered by peptide simulations [4–7]. Various methods have been proposed to overcome the conformational sampling problem during molecular simulations (reviewed in [8]). Energy barriers can be crossed effectively with simulated annealing techniques that open a large conformational space at high simulation temperatures and select low energy states by cooling the system down [9–11].
Conformations of some lower-size large-ring cyclodextrins derived from conformational search with molecular dynamics and principal component analysis
2012, Journal of Molecular StructureCitation Excerpt :We proceed now further by extending our study to six other cyclodextrins with degree of polymerization from 11 to 17, thus aiming at acquiring knowledge for the conformations of the LR-CDs for the whole range for degree of polymerization from 10 to 30. Principal component analysis (PCA), also called quasiharmonic analysis or essential dynamics method [60–62], is a very useful analysis tool for MD trajectories [63–72]. This statistical method allows trajectories obtained from MD simulations to be analyzed by reducing the degrees of freedom of the system to lower dimensions [66].