Search

Peptide materials have a wide array of functions, from tissue engineering and surface coatings to catalysis and sensing. Tuning the sequence of amino acids that comprise the peptide modulates peptide functionality, but a small increase in sequence length leads to a dramatic increase in the number of peptide candidates. Traditionally, peptide design is guided by human expertise and intuition and typically yields fewer than ten peptides per study, but these approaches are not easily scalable and are susceptible to human bias. Here we introduce a machine learning workflow-AI-expert-that combines Monte Carlo tree search and random forest with molecular dynamics simulations to develop a fully autonomous computational search engine to discover peptide sequences with high potential for self-assembly. We demonstrate the efficacy of the AI-expert to efficiently search large spaces of tripeptides and pentapeptides. The predictability of AI-expert performs on par or better than our human experts and suggests several non-intuitive sequences with high self-assembly propensity, outlining its potential to overcome human bias and accelerate peptide discovery.

The development of synthetic structures that mimic mechanical actuation in living matter such as autonomous translation and shape changes remains a grand challenge for materials science. In living systems the integration of supramolecular structures and covalent polymers contributes to the responsive behaviour of membranes, muscles and tendons, among others. Here we describe hybrid light-responsive soft materials composed of peptide amphiphile supramolecular polymers chemically bonded to spiropyran-based networks that expel water in response to visible light. The supramolecular polymers form a reversibly deformable and water-draining skeleton that mechanically reinforces the hybrid and can also be aligned by printing methods. The noncovalent skeleton embedded in the network thus enables faster bending and flattening actuation of objects, as well as longer steps during the light-driven crawling motion of macroscopic films. Our work suggests that hybrid bonding polymers, which integrate supramolecular assemblies and covalent networks, offer strategies for the bottom-up design of soft matter that mimics living organisms. Peptide amphiphile supramolecular polymers with a crosslinked spiropyran network respond to light by expelling water, enabling the fabrication of soft actuators or light-driven crawlers.

A large variety of functional self-assembled supramolecular nanostructures have been reported over recent decades.1 The experimental approach to these systems initially focused on the design of molecules for specific interactions that lead to discrete geometric structures.1–4 Recently, kinetics and mechanistic pathways of self-assembly have been investigated,6,7 but there remains a major gap in our understanding of internal conformational dynamics and their links to function. This challenge has been addressed through computational chemistry with the introduction of molecular dynamics (MD) simulations, which yield information on molecular fluctuations over time.5–7 Experimentally, it has been difficult to obtain analogous data with sub-nanometer spatial resolution. Thus, there is a need for experimental dynamics measurements, to confirm and guide computational efforts and to gain insight into the internal motion in supramolecular assemblies. Using site-directed spin labeling and electron paramagnetic resonance (EPR) spectroscopy, we measured conformational dynamics through the 6.7 nm cross-section of a self-assembled nanofiber in water and provide unique insight for the design of supramolecular functional materials.

Aggregates of charged amphiphilic molecules have been found to access a structure at elevated temperature that templates alignment of supramolecular fibrils over macroscopic scales. The thermal pathway leads to a lamellar plaque structure with fibrous texture that breaks upon cooling into large arrays of aligned nanoscale fibres and forms a strongly birefringent liquid. By manually dragging this liquid crystal from a pipette onto salty media, it is possible to extend this alignment over centimetres in noodle-shaped viscoelastic strings. Using this approach, the solution of supramolecular filaments can be mixed with cells at physiological temperatures to form monodomain gels of aligned cells and filaments. The nature of the self-assembly process and its biocompatibility would allow formation of cellular wires in situ that have any length and customized peptide compositions for use in biological applications.