(435e) Fast Prediction of Peptide-Surface Interaction without Massive Enhanced Sampling | AIChE

(435e) Fast Prediction of Peptide-Surface Interaction without Massive Enhanced Sampling


Qi, X. - Presenter, University of Washington
Pfaendtner, J., University of Washington
Functional peptides at the solid-liquid interface have demonstrated their versatility in guiding nanocrystal shape evolution, promoting biomimetic material design, and enabling applications such as biosensing. One of the crucial parameters in assessing their functionality is the binding free energy to a surface of the solid material. Although the binding energetics can be measured indirectly from experiments, the vast majority nowadays is reported through computational means thanks to the rapid development of computational power. Molecular dynamics (MD) simulation with metadynamics (MetaD) has been one of the most prevalent methods in estimating the peptide-surface interaction. However, for the next-generation goal of simulation-leading programmable material design, the combination MD and MetaD appears to be time and resource consuming because each free energy calculation is specific for a given peptide sequence.

We note from previous studies on peptide adsorption at an interface that, while the binding energetics largely depend on how accessible each amino acid (AA) residue is to the surface at the equilibrium, a large portion of computational effort goes into the exploration in the solution phase in MetaD sampling. To speed up the calculation, a generic relation between the equilibrium peptide configuration and the binding energetics needs to be established. Using our model system, a genetically engineered silica-binding peptide Car9 adsorbed on a quartz surface, we find that binding free energy of the entire peptide chain can be expressed as a weighted sum of the binding free energy of its consisting AA residue headgroups at the equilibrium height from the surface. We further validate our hypothesis against other Car9 mutants binding on quartz. Since the binding free energy of single AA on a surface requires much less computational effort and can be repeatedly utilized, our generic model is able to realize fast prediction of peptide-surface interaction without massive MetaD sampling and enable high throughput energetic scanning for simulation-leading programmable material design.

This work was supported by the US Department of Energy, Office of Science, Office of Basic Energy Sciences, as part of the Energy Frontier Research Centers program: CSSAS (The Center for the Science of Synthesis Across Scales) under Award Number DE-SC0019288.