High-Throughput Computational Screening of Solid-Binding Peptides.

Area of Science:

• Biomineralization and Nanomaterial Science
• Computational Chemistry and Peptide Design

Background:

• Solid-binding peptides (SBPs) are inspired by natural biomineralization for creating nanostructures.
• Designing SBPs with specific binding properties traditionally requires intensive methods like phage display.
• Existing computational validation methods for SBP sequences are often prohibitively expensive.

Purpose of the Study:

• To develop a rapid and cost-effective computational model for estimating SBP binding free energy to solid surfaces.
• To enable faster screening and design of SBPs for targeted applications.
• To reduce the computational cost associated with SBP sequence validation.

Main Methods:

• Developed a novel model to estimate binding free energy based on residue contributions from stable structure ensembles.
• Utilized statistical analysis of thermodynamically stable structures.
• Validated the model using five silica-binding peptides with varying affinities and lengths.

Main Results:

• The model accurately and robustly estimates binding free energy across diverse chemistries and binding strengths.
• Achieved a computational cost as low as 3% compared to traditional enhanced sampling methods.
• Demonstrated the model's effectiveness for silica-binding peptides.

Conclusions:

• The new model significantly accelerates the estimation of SBP-surface binding free energy.
• Offers a cost-effective alternative to existing validation techniques.
• Has high potential for integration into high-throughput screening and machine learning algorithms for SBP discovery.