Predicting helical hairpins from sequences by Monte Carlo simulations

The Puzzle of RNAs that Target Gene Promoters

“…The application of OPEP1.0‐MC simulations to a series of 30 peptides yields equilibrium conformations consistent with NMR data in most test cases 8–10. However, for four HTH models, it was found that the short‐range forces controlling the stereochemical properties of the amino acids were not always compatible with the equilibrium conformations found within the protein data bank 10. As a result, the energy function was adjusted by maximizing the energy gap between the native and misfolded conformations for the four training peptides8 and two de novo designed HTH models, termed Z38 and Z34C 10.…”

“…The weights w of the different terms in the OPEP energy function were first optimized on the NMR structures of four peptides with β hairpin, α‐helix, αβ, and ββα conformational preferences 8. The application of OPEP1.0‐MC simulations to a series of 30 peptides yields equilibrium conformations consistent with NMR data in most test cases 8–10. However, for four HTH models, it was found that the short‐range forces controlling the stereochemical properties of the amino acids were not always compatible with the equilibrium conformations found within the protein data bank 10.…”

“…A complete description of the OPEP‐Monte Carlo (MC) procedure has been given elsewhere 8–10. In brief, it uses a simplified chain representation with all backbone atoms included (i.e., N, H, C α , C, and O) and all side chains modeled by a bead with an appropriate van der Waals radius and geometry.…”

“…The transition time between the trial and current conformations is estimated from the diffusive motion of the atoms in the (ϕ, ψ) dihedral space. The resulting minimized conformation10 is accepted or rejected, using the classical Metropolis criterion 19…”

“…This balance explains why identical fragment sequences adopt distinct conformations within protein structures. To this end, we report on the application of a recent kinetic Monte Carlo–based procedure8–10 to determine the equilibrium structures of a test set of polypeptides with three‐stranded antiparallel beta‐sheet (β 3 ), α‐helical hairpin, and βαβ conformational preferences in solution. This method, which builds a structure for the target sequence by the global optimization of an effective potential, has already been applied to a series of model peptides, including β‐hairpins8 and three‐helix bundles 9.…”