Results 1  10
of
15
Roadmap Methods for Protein Folding
"... Abstract—Protein folding refers to the process whereby a protein assumes its intricate threedimensional shape. �is chapter reviews a class of methods for studying the folding process called roadmap methods. �e goal of these methods is not to predict the folded structure of a protein, but rather to ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
(Show Context)
Abstract—Protein folding refers to the process whereby a protein assumes its intricate threedimensional shape. �is chapter reviews a class of methods for studying the folding process called roadmap methods. �e goal of these methods is not to predict the folded structure of a protein, but rather to analyze the folding kinetics. It is assumed that the folded state is known. Roadmap methods build a graph representation of sampled conformations. By analyzing this graph one can predict structure formation order, the probability of folding, and get a coarse view of the energy landscape.
Particle Markov Chain Monte Carlo
, 2009
"... ... have emerged as the two main tools to sample from highdimensional probability distributions. Although asymptotic convergence of MCMC algorithms is ensured under weak assumptions, the performance of these latters is unreliable when the proposal distributions used to explore the space are poorly ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
... have emerged as the two main tools to sample from highdimensional probability distributions. Although asymptotic convergence of MCMC algorithms is ensured under weak assumptions, the performance of these latters is unreliable when the proposal distributions used to explore the space are poorly chosen and/or if highly correlated variables are updated independently. In this thesis we propose a new Monte Carlo framework in which we build efficient highdimensional proposal distributions using SMC methods. This allows us to design effective MCMC algorithms in complex scenarios where standard strategies fail. We demonstrate these algorithms on a number of example problems, including simulated tempering, nonlinear nonGaussian statespace model, and protein folding.
permission. Resampling Methods for Protein Structure Prediction
"... personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires pri ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific
Improving HighDimensional Bayesian Network Structure Learning by Exploiting Search Space Information
, 2006
"... Bayesian networks are frequently used to model statistical dependencies in data. Without prior knowledge of dependencies in the data, the structure of a Bayesian network is learned from the data. Bayesian network structure learning is commonly posed as an optimization problem where search is used to ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
Bayesian networks are frequently used to model statistical dependencies in data. Without prior knowledge of dependencies in the data, the structure of a Bayesian network is learned from the data. Bayesian network structure learning is commonly posed as an optimization problem where search is used to find structures that maximize a scoring function. Since the structure search space is superexponential in the number of variables in a network, heuristics are applied to constrain the search space of highdimensional networks. Greedy hill climbing is then applied in the reduced search space. The constrained search space of highdimensional networks contains many local maxima that greedy hill climbing cannot overcome. This issue has only been addressed by augmenting greedy search with TABU lists or random moves. This is not a holistic solution to the problem. By using a search algorithm that is global in nature, we are not confined to results in a particular region of the search space, like previous approaches. We present ModelBased Search (MBS) [1] applied to Bayesian network structure learning. MBS uses information gained during search to explore promising search space regions. Maintaining this search space information keeps a global view of the search task and helps find structures at higher maxima than greedy hill climbing. We show that MBS performs better than hill climbing in the MaxMin Parents and Children (MMPC) [30] search space and can find better highdimensional network structures than other leading structure learning algorithms. 1
Novel Heuristic Search Methods for Protein Folding and Identification of Folding Pathways
, 2006
"... Proteins form the very basis of life. If we were to open up any living cell, we would find, apart from DNA and RNA molecules whose primary role is to store genetic information, a large number of different proteins that comprise the cell itself (for example the cell membrane and organelles), as well ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
Proteins form the very basis of life. If we were to open up any living cell, we would find, apart from DNA and RNA molecules whose primary role is to store genetic information, a large number of different proteins that comprise the cell itself (for example the cell membrane and organelles), as well as a diverse set of enzymes that catalyze various metabolic reactions. If enzymes were absent, the cell would not be able to function, since a number of metabolic reactions would not be possible. Functions of proteins are the consequences of their functional 3D shape. Therefore, to control these versatile properties, we need to be able to predict the 3D shape of proteins; in other words, solve the protein folding problem. The prediction of a protein’s conformation from its aminoacid sequence is currently one of the most prominent problems in molecular biology, biochemistry and bioinformatics. In this thesis, we address the protein folding problem and the closelyrelated problem of identifying folding pathways. The leading research objective for this work was to design efficient heuristic search algorithms for these problems, to empirically
An extremal optimization search method for the protein folding problem: the gomodel example
 In: Proceedings of the 2007 GECCO conference companion on Genetic and evolutionary computation
, 2007
"... The protein folding problem consists of predicting the functional (native) structure of the protein given its linear sequence of amino acids. Despite extensive progress made in understanding the process of protein folding, this problem still remains extremely challenging. In this paper we introduce, ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
The protein folding problem consists of predicting the functional (native) structure of the protein given its linear sequence of amino acids. Despite extensive progress made in understanding the process of protein folding, this problem still remains extremely challenging. In this paper we introduce, implement and evaluate the Extremal Optimization method – a biologically inspired approach which has been applied very successfully to other optimization problems – for the protein folding problem using a widely studied Gōmodel of folding. Standard methods based on the variants of the Monte Carlo method have difficulty exploring lowenergy regions efficiently due to the ruggedness of the search landscapes. Most computational methods in the protein folding literature do not keep track of which interactions remain unsatisfied during the search. Instead, in this paper, we propose an adaptive metasearch method which ensures that unexplored promising parts of the search landscape are visited. This is achieved by implementing an adaptive Extremal Optimization metasearch that guides a standard Monte Carlo sampling. We demonstrate that our Extremal Optimization metasearch compares favorably with currently bestperforming Replica Exchange Monte Carlo method in reaching the native state for long proteins under the Gōmodel potential. Additionally, we show that our novel approach samples larger ensembles of nearnative structures by plotting parts of the energy landscape sampled during the search. Furthermore, we find that it scales well with the increasing sequence length. To our best knowledge this is the first application of Extremal Optimization to the protein folding problem.
Improved search for structure learning of large bayesian networks
"... The problem of Bayesian network structure learning is defined as an optimization problem over the space of all possible network structures. For lowdimensional data, optimal structure learning approaches exist. For highdimensional data, structure learning remains a significant challenge. Most commo ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
The problem of Bayesian network structure learning is defined as an optimization problem over the space of all possible network structures. For lowdimensional data, optimal structure learning approaches exist. For highdimensional data, structure learning remains a significant challenge. Most commonly, approaches to highdimensional structure learning employ a reduced search space and apply hill climbing methods to find highscoring network structures. But even the reduced search space contains many local optima so that local search methods are unable to find nearoptimal network structures. Instead of focusing on search space reduction, as most of the previous work in this area, we propose to replace the greedy search schemes with more effective search methods. We show that for highdimensional data the proposed search method finds significantly better structures than other leading approaches to structure learning. 1
A Probabilistic FragmentBased Protein Structure Prediction Algorithm
 PLOS One
, 2012
"... Conformational sampling is one of the bottlenecks in fragmentbased protein structure prediction approaches. They generally start with a coarsegrained optimization where mainchain atoms and centroids of side chains are considered, followed by a finegrained optimization with an allatom representat ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
Conformational sampling is one of the bottlenecks in fragmentbased protein structure prediction approaches. They generally start with a coarsegrained optimization where mainchain atoms and centroids of side chains are considered, followed by a finegrained optimization with an allatom representation of proteins. It is during this coarsegrained phase that fragmentbased methods sample intensely the conformational space. If the nativelike region is sampled more, the accuracy of the final allatom predictions may be improved accordingly. In this work we present EdaFold, a new method for fragmentbased protein structure prediction based on an Estimation of Distribution Algorithm. Fragmentbased approaches build protein models by assembling short fragments from known protein structures. Whereas the probability mass functions over the fragment libraries are uniform in the usual case, we propose an algorithm that learns from previously generated decoys and steers the search toward nativelike regions. A comparison with Rosetta AbInitio protocol shows that EdaFold is able to generate models with lower energies and to enhance the percentage of nearnative coarsegrained decoys on a benchmark of 20 proteins. The best coarsegrained models produced by both methods were refined into allatom models and used in molecular replacement. All atom decoys produced out of EdaFold’s decoy set reach high enough accuracy to solve the crystallographic phase problem by molecular replacement for some test proteins. EdaFold showed a higher success rate in molecular replacement when compared to Rosetta. Our study suggests that improving low resolution
Feature Space Resampling for Protein Conformational Search Short title: Protein Feature Space Resampling
"... De novo protein structure prediction requires location of the lowest energy state of the polypeptide chain among a vast set of possible conformations. Powerful approaches include conformational space annealing, in which search progressively focuses on the most promising regions of conformational spa ..."
Abstract
 Add to MetaCart
(Show Context)
De novo protein structure prediction requires location of the lowest energy state of the polypeptide chain among a vast set of possible conformations. Powerful approaches include conformational space annealing, in which search progressively focuses on the most promising regions of conformational space, and genetic algorithms, in which features of the best conformations thus far identified are recombined. We describe a new approach that combines the strengths of these two approaches. Protein conformations are projected onto a discrete feature space which includes backbone torsion angles, secondary structure, and beta pairings. For each of these there is one “native ” value: the one found in the native structure. We begin with a large number of conformations generated in independent Monte Carlo structure prediction trajectories from Rosetta. Native values for each feature are predicted from the frequencies of feature value occurrences and the energy distribution in conformations containing them. A second round of structure prediction trajectories are then guided by the predicted native feature distributions. We show that native features can be predicted at much higher than background rates, and that using the predicted feature distributions improves structure prediction in a benchmark of 28 proteins. Our approach allows generation of successful models by recombining nativelike parts of firstround conformations. The advantages of our approach are that features from many different input structures can be combined simultaneously without producing atomic clashes or otherwise physically unviable models, and that