Toward accurately modeling N-methylated cyclic peptides.

Area of Science:

• Computational chemistry
• Peptide science
• Drug discovery

Background:

• Cyclic peptides offer specific and potent targeting of protein surfaces.
• N-methylation is a strategy to optimize cyclic peptide pharmacokinetics and structures.
• Accurate modeling of N-methylated cyclic peptides is crucial for rational drug design.

Purpose of the Study:

• To evaluate molecular dynamics simulations and force fields for characterizing N-methylated cyclic peptide structures.
• To assess the performance of residue-specific force fields, particularly RSFF2, for N-methylated cyclic peptides.

Main Methods:

• Utilized molecular dynamics simulations with enhanced sampling techniques.
• Tested multiple simulation force fields, focusing on residue-specific force fields like RSFF2.
• Employed experimental structures as benchmarks for N-methylated cyclic peptides.

Main Results:

• The RSFF2 force field accurately reproduced experimental structures of N-methylated cyclic peptides.
• Accurate structure prediction was achieved when correct amide isomers were provided as initial configurations.
• The simulation approach demonstrated efficiency in characterizing structural ensembles.

Conclusions:

• Molecular dynamics simulations, particularly with RSFF2 and correct amide isomer information, can accurately predict N-methylated cyclic peptide structures.
• Further optimization of RSFF2 for predicting cis/trans amide isomers will enable *de novo* prediction of these molecules.
• This work strongly supports continued development of simulation methods for cyclic peptide drug design.