Discovery of Self-Assembling π-Conjugated Peptides by Active Learning-Directed Coarse-Grained Molecular Simulation

Area of Science:

• Materials Science
• Computational Chemistry
• Biotechnology

Background:

• Electronically active organic molecules are promising for energy applications.
• Self-assembled nanoaggregates from π-conjugated oligopeptides offer tunable optoelectronic properties.
• Understanding sequence-structure-function relationships in these nanoaggregates is crucial but poorly characterized.

Purpose of the Study:

• To efficiently identify molecules capable of assembling pseudo-1D nanoaggregates with optimal π-core stacking.
• To explore the DXXX-OPV3-XXXD oligopeptide family and identify top-performing tripeptide sequences.
• To establish new design rules for peptide-based nanoaggregate assembly.

Main Methods:

• Coarse-grained molecular dynamics simulations.
• Active learning protocol integrating deep representational learning and Bayesian optimization.
• Spectral clustering for analyzing assembly patterns.

Main Results:

• Identified superior tripeptide sequences within the DXXX-OPV3-XXXD family by simulating only 2.3% of the sequence space.
• Discovered new design principles governing the assembly of these oligopeptides.
• Predicted molecules with enhanced assembly properties compared to prior work.

Conclusions:

• Established a deeper understanding of DXXX-OPV3-XXXD nanoaggregate assembly.
• Identified promising candidates for experimental validation in energy harvesting and transport.
• Developed a versatile computational platform for designing other peptide-based soft materials.