Loading...
Please wait, while we are loading the content...
Similar Documents
All-atom folding of the trp-cage protein with an adpative parallel tempering method
| Content Provider | Semantic Scholar |
|---|---|
| Author | Schug, Alexander Wenzel, Wolfgang |
| Copyright Year | 2004 |
| Abstract | – Using the recently developed protein free-energy force field PFF01 we report the reproducible all-atom folding of the 20 amino acid trp-cage protein to within 2.0 Å backbone RMS deviation to the experimental structure with modest computational resources. We used an adapted version of the parallel tempering method as an inherently parallel stochastic optimization method. We find that near native structures dominate the low-energy spectrum of the final conformations and investigate the efficiency of the method as a function of the number of replicas in application to all-atom protein structure prediction. Protein structure prediction on the basis of the amino acid sequence alone remains one of the major outstanding challenges of theoretical biophysics [1–5]. In the post-genomic era, sequence information for proteins abounds, while structural and mechanistic information remains scarce. Theoretical methods for protein structure prediction may help to close the gap between the available sequences and structures and elucidate mechanisms of proteins that are difficult to handle experimentally (e.g., transmembrane proteins). With the development of reliable force fields [6,7] and robust simulation techniques [4,5,8], protein structure prediction may assist in the understanding and quantitative analysis of protein-protein or protein-ligand association [9, 10] at an atomistic level. While homology-based methods have demonstrated steady progress in the past decade [11], the assessment of atomistic de novo prediction strategies has been less favorable [2, 3, 12]. Atomistic simulations of the folding process remain confined to small peptides due to their large computational cost [13–15]. Based on the thermodynamic paradigm of protein folding [16], free-energy–based methods describe the native structure of the protein as the global optimum of a suitable free-energy force field. This approach is potentially much faster and more predictive than the costly simulation of the folding pathway, but will obviously sacrifice dynamical information. We have recently reported the rational development of a transferable all-atom free-energy force field (PFF01) [7] that correctly predicts the native structure of several proteins with 2060 amino acids, as the global minimum of the free-energy surface (FES). Reproducible folding could be demonstrated for the 20 amino acid trp-cage protein [4], the 40 amino acid headgroup of the HIV accessory protein [5] and the 36 amino acid headgroup of villin [8, 13, 17, 18]. An alternate free-energy force field was recently used to predict the structure of the B domain of staphylococcal protein A from first principles [6]. |
| File Format | PDF HTM / HTML |
| Language | English |
| Access Restriction | Open |
| Subject Keyword | Algorithmic efficiency Amino Acid Sequence Amino Acids Amino Acids, Branched-Chain Atom Computation Computational resource De novo transcriptome assembly Dynamical system Experiment Force field (chemistry) Gene prediction Gene regulatory network Global optimization Homologous Gene Homology (biology) Internet backbone Ligands Mathematical optimization Maxima and minima Molecular dynamics Parallel tempering Programming paradigm Protein structure prediction Protein, Organized by Structure Shin Megami Tensei: Persona 3 Simulation Stochastic optimization Thermodynamics VIL1 gene Vertebral column protein folding |
| Content Type | Text |
| Resource Type | Article |
