Organic Crystal Structure Prediction Research Articles

Machine-Learned Potentials by Active Learning from Organic Crystal Structure Prediction Landscapes.

A primary challenge in organic molecular crystal structure prediction (CSP) is accurately ranking the energies of potential structures. While high-level solid-state density functional theory (DFT) methods allow for mostly reliable discrimination of the low-energy structures, their high computational cost is problematic because of the need to evaluate tens to hundreds of thousands of trial crystal structures to fully explore typical crystal energy landscapes. Consequently, lower-cost but less accurate empirical force fields are often used, sometimes as the first stage of a hierarchical scheme involving multiple stages of increasingly accurate energy calculations. Machine-learned interatomic potentials (MLIPs), trained to reproduce the results of ab initio methods with computational costs close to those of force fields, can improve the efficiency of the CSP by reducing or eliminating the need for costly DFT calculations. Here, we investigate active learning methods for training MLIPs with CSP datasets. The combination of active learning with the well-developed sampling methods from CSP yields potentials in a highly automated workflow that are relevant over a wide range of the crystal packing space. To demonstrate these potentials, we illustrate efficiently reranking large, diverse crystal structure landscapes to near-DFT accuracy from force field-based CSP, improving the reliability of the final energy ranking. Furthermore, we demonstrate how these potentials can be extended to more accurately model structures far from lattice energy minima through additional on-the-fly training within Monte Carlo simulations.

Organic crystal structure prediction via coupled generative adversarial networks and graph convolutional networks

Organic crystal structures exert a profound impact on the physicochemical properties and biological effects of organic compounds. Quantum mechanics (QM)-based crystal structure predictions (CSPs) have somewhat alleviated the dilemma that experimental crystal structure investigations struggle to conduct complete polymorphism studies, but the high computing cost poses a challenge to its widespread application. The present study aims to construct DeepCSP, a feasible pure machine learning framework for minute-scale rapid organic CSP. Initially, based on 177,746 data entries from the Cambridge Crystal Structure Database, a generative adversarial network was built to conditionally generate trial crystal structures under selected feature constraints for the given molecule. Simultaneously, a graph convolutional attention network was used to predict the density of stable crystal structures for the input molecule. Subsequently, the distances between the predicted density and the definition-based calculated density would be considered to be the crystal structure screening and ranking basis, and finally, the density-based crystal structure ranking would be output. Two such distinct algorithms, performing the generation and ranking functionalities, respectively, collectively constitute the DeepCSP, which has demonstrated compelling performance in marketed drug validations, achieving an accuracy rate exceeding 80% and a hit rate surpassing 85%. Inspiringly, the computing speed of the pure machine learning methodology demonstrates the potential of artificial intelligence in advancing CSP research.

Accelerated Organic Crystal Structure Prediction with Genetic Algorithms and Machine Learning.

We present a high-throughput, end-to-end pipeline for organic crystal structure prediction (CSP)─the problem of identifying the stable crystal structures that will form from a given molecule based only on its molecular composition. Our tool uses neural network potentials to allow for efficient screening and structural relaxation of generated crystal candidates. Our pipeline consists of two distinct stages: random search, whereby crystal candidates are randomly generated and screened, and optimization, where a genetic algorithm (GA) optimizes this screened population. We assess the performance of each stage of our pipeline on 21 molecules taken from the Cambridge Crystallographic Data Centre's CSP blind tests. We show that random search alone yields matches for ≈50% of targets. We then validate the potential of our full pipeline, making use of the GA to optimize the root-mean-square deviation between crystal candidates and the experimentally derived structure. With this approach, we are able to find matches for ≈80% of candidates with 10-100 times smaller initial population sizes than when using random search. Lastly, we run our full pipeline with an ANI model that is trained on a small data set of molecules extracted from crystal structures in the Cambridge Structural Database, generating ≈60% of targets. By leveraging machine learning models trained to predict energies at the density functional theory level, our pipeline has the potential to approach the accuracy of ab initio methods and the efficiency of empirical force fields.

In silico and in vitro analysis of THCA synthase gene in Moroccan Cannabis sativa, L

Background: Cannabis ( Cannabis sativa, L.) is an archaic, most infamous plant with unique therapeutic characteristics, and great economic interest. The identification of its THCAS synthesizing enzyme was a great leap forward in Cannabis investigation. However, basic molecular biology for THCAS gene expression remains largely unstudied. In this paper, we explored the ability of bioinformatics resources in the design of molecular cloning protocol and prediction of three-dimensional structure of recombinant THCAS in Moroccan Cannabis sativa variety. Methods: We described conventional experiments performed in vitro. We used specific primers to isolate the nucleotide sequence of THCAS from the leaves of Cannabis, without including the native signal sequence.we prepared the gene by amors containing specific adapters to integrate the gene into the expression vector PpinkαHC, transforming E.coli bacteria, in order to express the gene in Pichia pasteri yeast. Results: As a result, we found high sequence similarity with THCAS and the mRNA precursor of the same gene as previously reported. In addition, we predicted the structure of the recombinant THCAS using the organic crystal structure prediction method, the amino acid sequence reference, and the artificial intelligence technology. Conclusions: In this study, we have successfuly cloned recombinant THCAS enzyme in E.Coli for further expression in host Pichia pastori.

In silico and in vitro analysis of THCA synthase gene in Moroccan Cannabis Sativa, L

Background: Cannabis (Cannabis sativa, L.) is an archaic, most infamous plant with unique therapeutic characteristics, and great economic interest. The identification of its THCAS synthesizing enzyme was a great leap forward in Cannabis investigation. However, basic molecular biology for THCAS gene expression remains largely unstudied. Methods: In this paper, we explored the ability of bioinformatics resources in the design of molecular cloning protocol and prediction of three-dimensional structure of recombinant THCAS in Moroccan Cannabis sativa variety. We further described conventional experiments performed in vitro. We used specific markers to isolate the nucleotide sequence of THCAS from the leaves of Cannabis, without including the native signal sequence. Results: As a result, we found high sequence similarity with THCAS and the mRNA precursor of the same gene as previously reported. In addition, we predicted the structure of the recombinant THCAS using the organic crystal structure prediction method, the amino acid sequence reference, and the artificial intelligence technology. Conclusions: This study paves the way to successfully express recombinant THCAS enzyme in hosts like Pichia pastori.

Influences of electrostatic models on organic crystal structure prediction – a case study of pentacene

Four electrostatic models including Charge, AMP, AMOEBA and QM-PFF models are applied to explore molecular packing of pentacene. AMP is a decisive factor to obtain the correct herringbone packing structures.

Data-efficient machine learning for molecular crystal structure prediction.

The combination of modern machine learning (ML) approaches with high-quality data from quantum mechanical (QM) calculations can yield models with an unrivalled accuracy/cost ratio. However, such methods are ultimately limited by the computational effort required to produce the reference data. In particular, reference calculations for periodic systems with many atoms can become prohibitively expensive for higher levels of theory. This trade-off is critical in the context of organic crystal structure prediction (CSP). Here, a data-efficient ML approach would be highly desirable, since screening a huge space of possible polymorphs in a narrow energy range requires the assessment of a large number of trial structures with high accuracy. In this contribution, we present tailored Δ-ML models that allow screening a wide range of crystal candidates while adequately describing the subtle interplay between intermolecular interactions such as H-bonding and many-body dispersion effects. This is achieved by enhancing a physics-based description of long-range interactions at the density functional tight binding (DFTB) level—for which an efficient implementation is available—with a short-range ML model trained on high-quality first-principles reference data. The presented workflow is broadly applicable to different molecular materials, without the need for a single periodic calculation at the reference level of theory. We show that this even allows the use of wavefunction methods in CSP.

Organic crystal structure prediction – from fundamental research to industrial application

Organic crystal structure prediction – from fundamental research to industrial application

Report on the sixth blind test of organic crystal structure prediction methods.

The sixth blind test of organic crystal structure prediction (CSP) methods has been held, with five target systems: a small nearly rigid molecule, a polymorphic former drug candidate, a chloride salt hydrate, a co-crystal and a bulky flexible molecule. This blind test has seen substantial growth in the number of participants, with the broad range of prediction methods giving a unique insight into the state of the art in the field. Significant progress has been seen in treating flexible molecules, usage of hierarchical approaches to ranking structures, the application of density-functional approximations, and the establishment of new workflows and `best practices' for performing CSP calculations. All of the targets, apart from a single potentially disordered Z' = 2 polymorph of the drug candidate, were predicted by at least one submission. Despite many remaining challenges, it is clear that CSP methods are becoming more applicable to a wider range of real systems, including salts, hydrates and larger flexible molecules. The results also highlight the potential for CSP calculations to complement and augment experimental studies of organic solid forms.

Sixth blind test of organic crystal-structure prediction methods

Over the past 15 years progress in predicting crystal structures of small organic molecules has been charted by a series of blind tests hosted by the Cambridge Crystallographic Data Centre. This letter announces a sixth blind test to take place between September 2014 and August 2015, giving details of the target systems and the revised procedure. We hope that as many methods as possible will be assessed and benchmarked in this new blind test.

ChemInform Abstract: Predicting Crystal Structures of Organic Compounds

AbstractReview: 53 refs.

Accurate Modeling of Organic Molecular Crystals by Dispersion-Corrected Density Functional Tight Binding (DFTB).

The ambitious goal of organic crystal structure prediction challenges theoretical methods regarding their accuracy and efficiency. Dispersion-corrected density functional theory (DFT-D) in principle is applicable, but the computational demands, for example, to compute a huge number of polymorphs, are too high. Here, we demonstrate that this task can be carried out by a dispersion-corrected density functional tight binding (DFTB) method. The semiempirical Hamiltonian with the D3 correction can accurately and efficiently model both solid- and gas-phase inter- and intramolecular interactions at a speed up of 2 orders of magnitude compared to DFT-D. The mean absolute deviations for interaction (lattice) energies for various databases are typically 2-3 kcal/mol (10-20%), that is, only about two times larger than those for DFT-D. For zero-point phonon energies, small deviations of <0.5 kcal/mol compared to DFT-D are obtained.

Predicting crystal structures of organic compounds

Currently, organic crystal structure prediction (CSP) methods are based on searching for the most thermodynamically stable crystal structure, making various approximations in evaluating the crystal energy. The most stable (global minimum) structure provides a prediction of an experimental crystal structure. However, depending on the specific molecule, there may be other structures which are very close in energy. In this case, the other structures on the crystal energy landscape may be polymorphs, components of static or dynamic disorder in observed structures, or there may be no route to nucleating and growing these structures. A major reason for performing CSP studies is as a complement to solid form screening to see which alternative packings to the known polymorphs are thermodynamically feasible.

Accurate Induction Energies for Small Organic Molecules: 1. Theory.

The induction energy often plays a very important role in determining the structure and properties of clusters of organic molecules, but only in recent years has an effort been made to include this energy in such calculations, notably in the field of organic crystal structure prediction. In this paper and the following one in this issue we provide ab initio methods suitable for the accurate inclusion of the induction energy for molecules containing as many as 30 atoms or so. These techniques are based on Symmetry-Adapted Perturbation Theory using Density Functional Theory [SAPT(DFT)] and use distributed polarizabilities computed using the recently developed density-fitting algorithm with constrained refinement. With this approach we are able to obtain induction models of varying complexity and study the effects of overlap and related numerical issues. Basis set effects on the exact and asymptotic induction energies are investigated, and the roles of higher-order induction energies and many-body effects are explored.