Training Neural Network Models Using Molecular Dynamics Simulation Results to Efficiently Predict Cyclic Hexapeptide Structural Ensembles

Molecular Dynamics (MD)-Derived Features for Canonical and Noncanonical Amino Acids

Machine learning (ML) models have become increasingly popular for predicting and designing structures and properties of peptides and proteins. These ML models typically use peptides and proteins containing only canonical amino acids as the training data. Consequently, these models struggle to make accurate predictions for peptides and proteins containing new amino acids that are absent in the training data set (e.g., noncanonical amino acids). One approach to improve the accuracy of the models is to collect more training data with the desired amino acids. However, this strategy is suboptimal as new data may not be easily attainable, and additional time is required to retrain the ML models. Alternatively, the extendibility of the ML models can be improved if the amino acid features used are representative and generalizable to the unseen amino acids. Herein, we develop amino acid features using molecular dynamics (MD) simulation results. Specifically, for a given amino acid, we perform MD simulation of its dipeptide to create features based on its backbone (ϕ, ψ) distributions and its electrostatic potentials. We demonstrate that these new features enable our ML models to more accurately predict the structural ensembles of cyclic peptides containing amino acids not present in the original training data set. For example, we build ML models to predict cyclic pentapeptide structures, with the training data set containing a library of 15 amino acids and the test data set containing the same 15-amino-acid library or an extended 50-amino-acid library. When using popular features such as Morgan fingerprints and MACCS keys to represent amino acids, the ML models achieve R2 = 0.963 for structural predictions of test cyclic pentapeptides containing the same 15-amino-acid library. However, these models' performances decrease significantly to R2 = 0.430 and R2 = 0.508, respectively, when tasked to predict the structures of cyclic pentapeptides containing a library of 50 amino acids. On the other hand, the model using our backbone (ϕ, ψ) features outperforms those using Morgan fingerprints and MACCS keys, with R2 = 0.700. Overall, instead of having to collect more training data, our new features enable predictions of peptide sequences containing amino acids not originally present in the training data set at the mere cost of performing new dipeptide simulations for the new amino acids.

Leveraging Cryptic Ligand Envelopes through Enhanced Molecular Simulations

Protein-bound ligands can adopt a range of different conformations, collectively defining a ligand envelope that has proven to be crucial for the design of potent and selective drugs. Yet, the cryptic nature of this ligand envelope makes it difficult to visualize, characterize, and ultimately exploit for drug design. Using enhanced molecular dynamics simulations, here, we provide a general framework to reconstruct the cryptic ligand envelope that is dynamically accessible by protein-bound small molecules in solution. We apply this approach to quantify hidden conformational heterogeneity in structurally complex ligands including the marine natural product plitidepsin. The computed conformational heterogeneity expands the small-molecule footprint beyond that typically observed in experiments, also revealing key thermodynamic and kinetic properties of single ligand-target interactions. The model agrees quantitatively with solution NMR, X-ray crystallography, and biochemical measurements, showcasing a versatile strategy to integrate receptor-bound ligand conformational ensembles in molecular design.

The structural and functional impacts of rationally designed cyclic peptides on self-assembly-mediated functionality

Compared with their linear counterparts, cyclic peptides, characterized by their unique topologies, offer superior stability and enhanced functionality. In this review article, the rational design of cyclic peptide primary structures and their significant influence on self-assembly processes and functional capabilities are comprehensively reviewed. We emphasize how strategically modifying amino acid sequences and ring sizes critically dictate the formation and properties of peptide nanotubes (PNTs) and complex assemblies, such as rotaxanes. Adjusting the number of amino acid residues and side chains allows researchers to tailor the diameter, surface properties, and functions of PNTs precisely. In addition, we discuss the complex host-guest chemistry of cyclic peptides and their ability to form rotaxanes, highlighting their potential in the development of mechanically interlocked structures with novel functionalities. Moreover, the critical role of computational methods for accurately predicting the solution structures of cyclic peptides is also highlighted, as it enables the design of novel peptides with tailored properties for a range of applications. These insights set the stage for groundbreaking advances in nanotechnology, drug delivery, and materials science, driven by the strategic design of cyclic peptide primary structures.

Assessing the Performance of Peptide Force Fields for Modeling the Solution Structural Ensembles of Cyclic Peptides

Molecular dynamics simulation is a powerful tool for characterizing the solution structural ensembles of cyclic peptides. However, the ability of simulation to recapitulate experimental results and make accurate predictions largely depends on the force fields used. In our work here, we evaluate the performance of seven state-of-the-art force fields in recapitulating the experimental NMR results in water of 12 benchmark cyclic peptides, consisting of 6 cyclic pentapeptides, 4 cyclic hexapeptides, and 2 cyclic heptapeptides. The results show that RSFF2+TIP3P, RSFF2C+TIP3P, and Amber14SB+TIP3P exhibit similar and the best performance, all recapitulating the NMR-derived structure information on 10 cyclic peptides. Amber19SB+OPC successfully recapitulates the NMR-derived structure information on 8 cyclic peptides. In contrast, OPLS-AA/M+TIP4P, Amber03+TIP3P, and Amber14SBonlysc+GB-neck2 could only recapitulate the NMR-derived structure information on 5 cyclic peptides, the majority of which are not well-structured.

Molecular Dynamics Simulations of Polyelectrolyte Complexes

Polyelectrolyte complexes (PECs) are currently of great interest due to their applications toward developing new adaptive materials and their relevance in membraneless organelles. These complexes emerge during phase separation when oppositely charged polymers are mixed in aqueous media. Peptide-based PECs are particularly useful toward developing new drug delivery methods due to their inherent biocompatibility. The underlying peptide sequence can be tuned to optimize specific material properties of the complex, such as interfacial tension and viscosity. Given their applicability, it would be advantageous to understand the underlying sequence-dependent phase behavior of oppositely charged peptides. Here, we report microsecond molecular dynamic simulations to characterize the effect of hydrophobicity on the sequence-dependent peptide conformation for model polypeptide sequences that were previously reported by Tabandeh et al. These sequences are designed with alternating chirality of the peptide backbone. We present microsecond simulations of six oppositely charged peptide pairs, characterizing the sequence-dependent effect on peptide size, degree of hydrogen bonding, secondary structure, and conformation. This analysis recapitulates sensible trends in peptide conformation and degree of hydrogen bonding, consistent with experimentally reported results. Ramachandran plots reveal that backbone conformation at the single amino acid level is highly influenced by the neighboring sequence in the chain. These results give insight into how subtle changes in hydrophobic side chain size and chirality influence the strength of hydrogen bonding between the chains and, ultimately, the secondary structure. Furthermore, principal component analysis reveals that the minimum energy structures may be subtly modulated by the underlying sequence.