Mimicking the folding pathway to improve homology-free protein structure prediction

Abstract

Since the demonstration that the sequence of a protein encodes its structure, the prediction of structure from sequence remains an outstanding problem that impacts numerous scientific disciplines, including many genome projects. By iteratively fixing secondary structure assignments of residues during Monte Carlo simulations of folding, our coarse-grained model without information concerning homology or explicit side chains can outperform current homologybased secondary structure prediction methods for many proteins. The computationally rapid algorithm using only single ( , ) dihedral angle moves also generates tertiary structures of accuracy comparable with existing all-atom methods for many small proteins, particularly those with low homology. Hence, given appropriate search strategies and scoring functions, reduced representations can be used for accurately predicting secondary structure and providing 3D structures, thereby increasing the size of proteins approachable by homology-free methods and the accuracy of template methods that depend on a high-quality input secondary structure.protein folding ͉ secondary structure prediction ͉ tertiary structure prediction ͉ iterative fixing ͉ statistical potential T he protein folding process is integral to multiple cellular processes, and errors can result in amyloidgenic diseases. The structure of a protein affords a window on its function, and the huge growth in the number of sequenced genomes provides codes for an enormous number of new proteins with unknown functions (1), a number far exceeding experimental capabilities and requiring fast throughput theoretical methods for deducing protein structure from sequence. To this end, great progress in predicting structure has emerged by using homology-based methods (2).However, the goal of predicting structure and pathways beginning only from the sequence remains an elusive goal. Furthermore, methods for secondary and tertiary (2°and 3°) structure prediction, although often quite accurate, can fail for lack of sufficient sequences that are homologous to the target sequence. Even if a multiple sequence alignment (MSA) exists, e.g., as generated by using PSI-BLAST (3), the alignment may diminish any structural propensity that is specific to the target sequence (in its 3°context) in favor of the consensus of the alignment. This disadvantage can adversely affect 3°structure prediction because the homologybased 2°structure prediction and MSA generally serve as crucial inputs.The reliance on homology also precludes identifying the underlying physiochemical principles that govern protein folding, including determining the minimal information and model of protein structure that are required for accurate structure prediction. This inadequacy arises from the failure of many 2°structure prediction methods (4, 5) to incorporate 3°context explicitly. Context dependence can overrule local biases (6-8), and its neglect has limited 2°s tructure accuracy to Ϸ80% for decades (9). Previous attempts to improve 2°structure predictions by inc...

Even with nonnative interactions, the updated folding transition states of the homologs Proteins G & L are extensive and similar

“…We conducted folding simulations using TerItFix (17,23,25,26,29), our homology-free, C β -level folding program that uses realistic sampling of the Ramachandran dihedral angles and authentic backbone H-bonding. Its Monte Carlo (MC) search strategy uses the principle of sequential stabilization to iteratively promote the formation of tertiary contacts and H-bonds across multiple rounds of folding.…”

“…This significant discrepancy is due in part to the presence of nonnative hairpin structures in the TSEs of the two naturally occurring proteins (15,16). These nonnative turns are likewise found in silico using our TerItFix folding algorithm (23)(24)(25)(26), which uses sequencedependent Ramachandran maps to predict native structures and folding pathways. Moreover, our experimental TSE for NuG2b proves to be in partial agreement with all-atom simulations (21).…”

Elucidation of Amyloid β-Protein Oligomerization Mechanisms: Discrete Molecular Dynamics Study

“…Intermediate-resolution models, such as the four-bead and united-atom models with backbone hydrogen bonding, which require knowledge only of the protein sequence, have been found to yield promising results when combined with DMD 22–26 or Monte Carlo dynamics. 27,28 …”

De novo prediction of protein folding pathways and structure using the principle of sequential stabilization

“…The principle of SS is implemented by using the statistics of folding trajectories garnered from prior rounds of simulation to bias the subsequent sampling of backbone dihedral angles (8) and the energies of tertiary contacts and hydrogen bonds. The approach combines simple backbone torsional ϕ, ψ moves, a polypeptide chain with no side chains beyond C β carbons, and multiple rounds of simulation with the progressive learning and building of 3°motifs through constraints imposed by data from prior rounds.…”