Heuristic energy-based cyclic peptide design.

Area of Science:

• Computational chemistry
• Drug discovery
• Peptide science

Background:

• Rational computational design is vital for discovering new drugs and therapeutic agents.
• Meso-scale cyclic peptides (7-40 amino acids) offer conformational rigidity, binding specificity, and resistance to degradation.
• De novo design of cyclic peptides, especially using non-canonical amino acids, is a promising strategy due to the scarcity of natural cyclic peptides.

Purpose of the Study:

• To develop an efficient computational pipeline, CyclicChamp, for the de novo design of cyclic peptides.
• To enable conformational sampling for larger cyclic peptides (15-24 residues) that were previously computationally intractable.
• To identify stable cyclic peptide designs through advanced molecular dynamics simulations.

Main Methods:

• Developed the CyclicChamp pipeline, converting cyclic constraints into an error function.
• Employed a variant of simulated annealing for efficient conformational sampling of peptide backbones.
• Utilized microsecond-length and replica exchange molecular dynamics simulations to assess kinetic and thermodynamic stability.

Main Results:

• CyclicChamp significantly accelerates conformational sampling for small macrocycles and addresses high-dimensionality challenges in larger designs.
• Conformational sampling became tractable for 15- to 24-residue cyclic peptides.
• Identified three 15-residue, one 20-residue, and one 24-residue cyclic peptide designs with significant kinetic and thermodynamic stability.

Conclusions:

• CyclicChamp provides a powerful and efficient method for designing novel meso-scale cyclic peptides.
• The developed computational approach facilitates the design of larger cyclic peptides, expanding possibilities in therapeutic agent discovery.
• The identified stable cyclic peptide candidates warrant further investigation for their potential therapeutic applications.