1,721,253 research outputs found
Dataset in support of the thesis 'Characterising the energy landscapes of molecular organic crystals'
No full textContains computational data described in the thesis, including:
- Threshold sampling with supercells and non-supercells, parallel trajectories, and convergence monitoring
- Transferable delta-ML potentials trained on CSP structures and re-ranked landscapes
- Geometry optimisation results using MACE potentials trained from CSP structures
For results that were published the computational data is stored with the corresponding publication record (see the linked datasets).
</span
Characterising the energy landscapes of molecular organic crystals
No full textThis thesis describes the development of methods for exploring, characterising, and fitting the energy landscapes of molecular crystals. A precise and comprehensive understanding of the energy landscapes of molecular crystals is essential for accurately predicting observed crystal structures and their behaviour. This is emphasised early through describing our efforts to predict the crystal structure of target XXVII from the recent blind test. Thereafter, using threshold Monte Carlo simulations and empirical force fields, we investigate sampling crystal landscapes of different compounds revealing key insights into general trends. In particular, we implement and test adaptive sampling based on monitoring the number of unique, energy minimised structures, finding considerable improvement over fixed sampling, especially at higher energies. Additionally, having noted that low energy connections can be converged with relative ease by the threshold algorithm, a method for reducing overprediction in organic crystal structure prediction is developed wherein low energy connections around predicted structures are sampled and then the landscape is clustered to the minimum energy structure in each of the resulting energy basins. This approach is tested on predicted landscapes for rigid and flexible molecules, with the results showing it can considerably reduce overprediction without discarding matches to experimentally observed structures. In the second half of the thesis studies into efficiently fitting the energy landscapes of organic crystals using machine learning potentials are detailed. These involve active learning from low-level predicted landscapes as well as on-the-fly training within Monte Carlo simulations. Potentials trained with these methods on a variety of compounds are shown to achieve excellent accuracy relative to high-level quantum chemistry calculations while maintaining low computational costs. Further case studies investigate extending these potentials to improve transferability between compounds and for achieving high-level geometry optimisations
Computational data for journal publication: "Reducing overprediction of molecular crystal structures via threshold clustering"
No full textComputational data related to landscapes of predicted crystal structures reported in: Reducing overprediction of molecular crystal structures via threshold clustering</span
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Computational data for: "Machine learned potentials by active learning from organic crystal structure prediction landscapes'
No full textComputational data related to landscapes of predicted crystal structures and neural network potentials reported in publication 'Machine Learned Potentials by Active Learning from Organic Crystal Structure Prediction Landscapes'.
in press Journal of Physical Chemistry A</span
Machine learned potentials by active learning from organic crystal structure prediction landscapes
No full textA 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.</p
Dataset: CSP-generated crystal structures of 1,000+ rigid organic molecules
No full textThis dataset supports the publication:
AUTHORS: Christopher R. Taylor, Patrick W. V. Butler, Graeme M. Day
TITLE: Predictive crystallography at scale: mapping, validating, and learning from 1,000 crystal energy landscapes
JOURNAL: Faraday Discussions
A consolidated dataset of crystal structure predictions (CSPs) for 1007 unique rigid, organic molecules with observed crystal structures in the Cambridge Structural Database (CSD). Each CSP is described by a "landscape" of hypothetical crystal structures, ranked in terms of their lattice energy; this dataset includes both the crystal structures themselves and their energy rankings.
This dataset also includes two machine-learning-derived models to improve the energy ranking of crystal structures on their respective landscapes; one a committee neural-network potential (NNP) to correct energies of fixed structures, the other a message-passing neural-network (MACE) model used to re-optimise particularly difficult crystal structures.</span
Predictive crystallography at scale: mapping, validating, and learning from 1,000 crystal energy landscapes
No full textComputational crystal structure prediction (CSP) is an increasingly powerful technique in materials discovery, due to its ability to reveal trends and permit insight across the possibility space of crystal structures of a candidate molecule, beyond simply the observed structure(s). In this work, we demonstrate the reliability and scalability of CSP methods for small, rigid organic molecules by performing in-depth CSP investigations for over 1000 such compounds, the largest survey of its kind to-date. We show that this highly-efficient force-field-based CSP approach is superbly predictive, locating 99.4\% of observed experimental structures, and ranking a large majority of these (74\%) as among the most stable possible structures (to within uncertainty due to thermal effects). We present two examples of insights such large predicted datasets can permit, examining the space group preferences of organic molecular crystals and rationalising empirical rules concerning the spontaneous resolution of chiral molecules. Finally, we exploit this large and diverse dataset for developing transferable machine-learned energy potentials for the organic solid state, training a neural network lattice energy correction to force field energies that offers substantial improvements to the already impressive energy rankings, and a MACE equivariant message-passing neural network for crystal structure reoptimisation. We conclude that the excellent performance and reliability of the CSP workflow enables the creation of very large datasets of broad utility and explanatory power in materials design
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
- …
