Mapping and classifying molecules from a high-throughput structural database.

High-throughput computational materials design promises to greatly accelerate the process of discovering new materials and compounds, and of optimizing their properties. The large databases of structures and properties that result from computational searches, as well as the agglomeration of data of heterogeneous provenance leads to considerable challenges when it comes to navigating the database, representing its structure at a glance, understanding structure-property relations, eliminating duplicates and identifying inconsistencies. Here we present a case study, based on a data set of conformers of amino acids and dipeptides, of how machine-learning techniques can help addressing these issues. We will exploit a recently-developed strategy to define a metric between structures, and use it as the basis of both clustering and dimensionality reduction techniques-showing how these can help reveal structure-property relations, identify outliers and inconsistent structures, and rationalise how perturbations (e.g. binding of ions to the molecule) affect the stability of different conformers.

Assessing the Accuracy of Across-the-Scale Methods for Predicting Carbohydrate Conformational Energies for the Examples of Glucose and α-Maltose.

A big hurdle when entering the field of carbohydrate research stems from the complications in the analytical and computational treatment. In effect, this extremely important class of biomolecules remains underinvestigated when compared, for example, with the maturity of genomics and proteomics research. On the theory side, the commonly used empirical methods suffer from an insufficient amount of high-quality experimental data against which they can be thoroughly validated. In order to provide a pivotal point for an ascent of accurate carbohydrate simulations, we present here a structure/energy benchmark set of diverse glucose (in three isomeric forms) and α-maltose conformations at the coupled-cluster level as well as an assessment of commonly used energy functions. We observe that empirical force fields and semiempirical approaches, on average, do not reproduce accurately the reference energy hierarchies. While the force fields maintain accuracy for the low-energy structures, the semiempirical methods perform unsatisfactory for the entire range. On the contrary, density-functional approximations reproduce the reference energy hierarchies with better than chemical accuracy already at the generalized gradient approximation level (GGA). Particularly, the novel meta-GGA functional SCAN provides the accuracy of more expensive hybrid functionals at fraction of their computational cost. In conclusion, we advocate for electronic-structure theory methods to become the routine tool for simulations of carbohydrates.