Search

Collagen is the most abundant structural protein in humans, with dozens of sequence variants accounting for over 30% of the protein in an animal body. The fibrillar and hierarchical arrangements of collagen are critical in providing mechanical properties with high strength and toughness. Due to this ubiquitous role in human tissues, collagen-based biomaterials are commonly used for tissue repairs and regeneration, requiring chemical and thermal stability over a range of temperatures during materials preparation ex vivo and subsequent utility in vivo. Collagen unfolds from a triple helix to a random coil structure during a temperature interval in which the midpoint or Tm is used as a measure to evaluate the thermal stability of the molecules. However, finding a robust framework to facilitate the design of a specific collagen sequence to yield a specific Tm remains a challenge, including using conventional molecular dynamics modeling. Here we propose a de novo framework to provide a model that outputs the Tm values of input collagen sequences by incorporating deep learning trained on a large data set of collagen sequences and corresponding Tm values. By using this framework, we are able to quickly evaluate how mutations and order in the primary sequence affect the stability of collagen triple helices. Specifically, we confirm that mutations to glycines, mutations in the middle of a sequence, and short sequence lengths cause the greatest drop in Tm values.

We propose a simple general framework to predict folding, native states, energy barriers, protein unfolding, as well as mutation induced diseases and other protein structural analyses. The model should not be considered as an alternative to classical approaches (Molecular Dynamics or Monte Carlo) because it neglects low scale details and rather focuses on global features of proteins and structural information. We aim at the description of phenomena that are out of the range of classical molecular modeling approaches due to the large computational cost: multimolecular interactions, cyclic behavior under variable external interactions, and similar. To demonstrate the effectiveness of the approach in a real case, we focus on the folding and unfolding behavior of tropoelastin and its mutations. Specifically, we derive a discrete mechanical model whose structure is deduced based on a coarse graining approach that allows us to group the amino acids sequence in a smaller number of `equivalent' masses. Nearest neighbor energy terms are then introduced to reproduce the interaction of such amino acid groups. Nearest and non-nearest neighbor energy terms, inter and intra functional blocks are phenomenologically added in the form of Morse potentials. As we show, the resulting system reproduces important properties of the folding-unfolding mechanical response, including the monotonic and cyclic force-elongation behavior, representing a physiologically important information for elastin. The comparison with the experimental behavior of mutated tropoelastin confirms the predictivity of the model. STATEMENT OF SIGNIFICANCE: Classical approaches to the study of phenomena at the molecular scale such as Molecular Dynamics (MD) represent an incredible tool to unveil mechanical and conformational properties of macromolecules, in particular for biological and medical applications. On the other hand, due to the computational cost, the time and spatial scales are limited. Focusing of the re

Fracture is a catastrophic and complex process that involves various time and length scales. Scientists have devoted vast efforts toward understanding the underlying mechanisms for centuries, with much work left in terms of predictability of models and fundamental understanding. To this end, we present a machine-learning approach to predict fracture processes connecting molecular simulation into a physics-based artificial intelligence (AI) multiscale model. Our model exhibits predictive power not only regarding the computed fracture patterns but also for fracture toughness—the resistance of cracks to grow. The novel AI-based fracture predictor can also deal with complex loading conditions, here examined for both mode I (tensile) and mode II (shear). These results underscore the excellent predictive power of our model. Potential applications include the design of novel types of high-performance materials, composites design, surface coatings, or innovative bio-inspired structures.

© 2017 Elsevier Ltd The nanometer-sized plate-like morphology of bone mineral is necessary for proper bone mechanics and physiology. However, mechanisms regulating the morphology of these mineral nanocrystals remain unclear. The dominant hypothesis attributes the size and shape regulation to organic-mineral interactions. Here, we present data supporting the hypothesis that physicochemical effects of carbonate integration within the apatite lattice control the morphology, size, and mechanics of bioapatite mineral crystals. Carbonated apatites synthesized in the absence of organic molecules presented plate-like morphologies and nanoscale crystallite dimensions. Experimentally-determined crystallite size, lattice spacing, solubility and atomic order were modified by carbonate concentration. Molecular dynamics (MD) simulations and density functional theory (DFT) calculations predicted changes in surface energy and elastic moduli with carbonate concentration. Combining these results with a scaling law predicted the experimentally observed scaling of size and energetics with carbonate concentration. The experiments and models describe a clear mechanism by which crystal dimensions are controlled by carbonate substitution. Furthermore, the results demonstrate that carbonate substitution is sufficient to drive the formation of bone-like crystallites. This new understanding points to pathways for biomimetic synthesis of novel, nanostructured biomaterials.

Fire has fascinated humankind since the prehistoric era. Rooted in the interactions between sound and flames, here we report a method to use fire for a variety of purposes, including sonification, art, and the design and manufacturing nature-inspired materials. We present a method to sonify fire, thereby offering a translation from the silent nature of flames, to represent audible information and to generate de novo flame images. To realize material specimen derived from fire, we use the autoencoder to generate image stacks to yield continuous 3D geometries that are manufactured using 3D printing. This represents the first generation of nature-inspired materials from fire and can be a platform to be used for other natural phenomena in the quest for de novo architectures, geometries, and design ideas, thus creating additional directions in artistic and scientific research through the creative manipulation of data with structural similarities across fields.

Wellcome Trust (Grant WT097347MA), NIH (Grants TUFTS-5U01EB014976 and WUSTL-5U01EB016422)

Elastin provides elastic tissues with resilience through stretch and recoil cycles, and is primarily made of itsextensively cross-linked monomer, tropoelastin. Here, we leverage the recently published full atomistic modelof tropoelastin to assess how allysine modifications, which are essential to cross-linking, contribute to thedynamics and structural changes that occur in tropoelastin in the context of elastin assembly. We used replicaexchange molecular dynamics to generate structural ensembles of allysine containing tropoelastin. Weconducted principal component analysis on these ensembles and found that the molecule departs from thecanonical structural ensemble. Furthermore, we showed that, while the canonical scissors-twist movementwas retained, new movements emerged that deviated from those of the wild type protein, providing evidencefor the involvement of a variety of molecular motions in elastin assembly. Additionally, we highlighted secondary structural changes and linked these perturbations to the longevity of specific salt bridges. Wepropose a model where allysines in tropoelastin contribute to hierarchical elastin assembly through global andlocal perturbations to molecular structure and dynamics., National Science Foundation (U.S.) (Grant ACI-1053575), United States. Office of Naval Research. Defense University Research Instrumentation Program (Grant N00014-17-1-2320), United States. Office of Naval Research (Grant N000141612333), National Institutes of Health (U.S.) (Grant 5U01EB014976)

The development of rational techniques to discover new mechanically relevant proteins for use in variety of applications ranging from mechanics, agriculture to biotechnology remains an outstanding nanomechanical design problem. The key barrier is to design a sequence to fold into a predictable structure to achieve a certain material function. Focused on alpha-helical proteins (as found in skin, hair, and many other mechanically relevant protein materials), we report a Multi-scale Neighborhood-based Neural Network (MNNN) model to learn how a specific amino acid sequence folds into a protein structure. The algorithm predicts the protein structure without using a template or co-evolutional information at a maximum error of 2.1 Å. We find that the prediction accuracy is higher than other models and the prediction consumes less than six orders of magnitude time than ab initio folding methods. We demonstrate that MNNN can predict the structure of an unknown protein that agrees with experiments, and our model hence shows a great advantage in the rational design of de novo proteins. Keywords: Protein; artificial intelligence; machine learning; deep neural networks; folding; structure prediction; computation