DART: Deep Learning Enabled Topological Interaction Model for Energy Prediction of Metal Clusters and its Application in Identifying Unique Low Energy Isomers

<div><div><div><p>Recently, Machine Learning (ML) has proven to yield fast and accurate predictions of chemical properties to accelerate the discovery of novel molecules and materials. The majority of the work is on organic molecules, and much more work needs to be done for inorganic molecules, especially clusters. In the present work, we introduce a simple Topological Atomic Descriptor called TAD, which encodes chemical environment information of each atom in the cluster. TAD is a simple and interpretable descriptor where each value represents the atom count in three shells. We also introduce the DART, Deep Learning Enabled Topological Interaction model, which uses TAD as a feature vector to predict energies of metal clusters, in our case Gallium clusters with size ranging from 31 to 70 atoms. DART model is designed based on the principle that energy is a function of atomic interactions and allows us to model these complex atomic interactions to predict the energy. We further introduce a new dataset called GNC_31-70, which comprises structures and DFT optimized energies of Gallium clusters with sizes ranging from 31 to 70 atoms. We show how DART can be used to accelerate the identification of ground-state structures without geometry optimization. Albeit using topological descriptor, DART achieves MAE of 3.59 kcal/mol (0.15 eV) on testset. We also show that our model can distinguish core and surface atoms in the Ga-70 cluster, which the model has never encountered earlier. Finally, we demonstrate the transferability of DART model by predicting energies for about 6k unseen configurations picked up from Molecular Dynamics (MD) data for three cluster sizes (46, 57, and 60) within seconds. The DART model was able to reduce the load on DFT optimizations while identifying unique low energy structures from MD data.</p></div></div></div>

DART: Deep Learning Enabled Topological Interaction Model for Energy Prediction of Metal Clusters and its Application in Identifying Unique Low Energy Isomers

<div><div><div><p>Recently, Machine Learning (ML) has proven to yield fast and accurate predictions of chemical properties to accelerate the discovery of novel molecules and materials. The majority of the work is on organic molecules, and much more work needs to be done for inorganic molecules, especially clusters. In the present work, we introduce a simple Topological Atomic Descriptor called TAD, which encodes chemical environment information of each atom in the cluster. TAD is a simple and interpretable descriptor where each value represents the atom count in three shells. We also introduce the DART, Deep Learning Enabled Topological Interaction model, which uses TAD as a feature vector to predict energies of metal clusters, in our case Gallium clusters with size ranging from 31 to 70 atoms. DART model is designed based on the principle that energy is a function of atomic interactions and allows us to model these complex atomic interactions to predict the energy. We further introduce a new dataset called GNC_31-70, which comprises structures and DFT optimized energies of Gallium clusters with sizes ranging from 31 to 70 atoms. We show how DART can be used to accelerate the identification of ground-state structures without geometry optimization. Albeit using topological descriptor, DART achieves MAE of 3.59 kcal/mol (0.15 eV) on testset. We also show that our model can distinguish core and surface atoms in the Ga-70 cluster, which the model has never encountered earlier. Finally, we demonstrate the transferability of DART model by predicting energies for about 6k unseen configurations picked up from Molecular Dynamics (MD) data for three cluster sizes (46, 57, and 60) within seconds. The DART model was able to reduce the load on DFT optimizations while identifying unique low energy structures from MD data.</p></div></div></div>

DART: deep learning enabled topological interaction model for energy prediction of metal clusters and its application in identifying unique low energy isomers

We introduce a simple topological atomic descriptor, TAD, and a deep learning enabled topological interaction model (DART) for predicting energies of metal clusters for efficient identification of unique clusters.

Epigenetic Regulation of WNT3A Enhancer during Regeneration of Injured Cortical Neurons

Traumatic brain injury is known to reprogram the epigenome. Chromatin immunoprecipitation-sequencing of histone H3 lysine 27 acetylation (H3K27ac) and tri-methylation of histone H3 at lysine 4 (H3K4me3) marks was performed to address the transcriptional regulation of candidate regeneration-associated genes. In this study, we identify a novel enhancer region for induced WNT3A transcription during regeneration of injured cortical neurons. We further demonstrated an increased mono-methylation of histone H3 at lysine 4 (H3K4me1) modification at this enhancer concomitant with a topological interaction between sub-regions of this enhancer and with promoter of WNT3A gene. Together, this study reports a novel mechanism for WNT3A gene transcription and reveals a potential therapeutic intervention for neuronal regeneration.

Hamiltonian formalism of topologically massive electrodynamics

We consider the four-dimensional topologically massive electrodynamics in which a gauge field interacts with rank two antisymmetric tensor field through a topological interaction. The photon becomes massive by eating the rank two tensor field, which is dual to the Higgs mechanism. We explicitly demonstrate the nature of the mechanism by performing a canonical analysis of the theory and discuss various aspects of it.

Topological interaction of electromagnetism and torsion

A fully gauge-invariant topological coupling of torsion to the electromagnetic field is examined. It is shown, while the gravitational fields are unaltered, torsion may serve as a source for electromagnetism and electromagnetism may serve as a source for torsion. Unlike most couplings, this gives rise to conservation of charge, no magnetic monopoles, and is in agreement with the principle of equivalence.

Disrupting Pitx1 regulatory topology results in overtly normal limb development

Gene expression patterns during development are orchestrated in part by thousands of distant-acting transcriptional enhancers. However, identifying enhancers that are essential for expression of their target genes has proven challenging. Genetic perturbation of individual enhancers in some cases results in profound molecular and developmental phenotypes, but in mild or no phenotypes in others. Topological maps of long-range regulatory interactions may provide the means to identify enhancers critical for developmental gene expression. Here, we leveraged chromatin topology to characterize and disrupt the major promoter-enhancer interaction for Pitx1, which is essential for hindlimb development. We found that Pitx1 primarily interacts with a single distal enhancer in the hindlimb. Using genome editing, we deleted this enhancer in the mouse. Although loss of the enhancer completely disrupts the predominant topological interaction in the Pitx1 locus, Pitx1 expression in the hindlimb is only reduced by ~14%, with no apparent changes in spatial distribution or evidence of regulatory compensation. Pitx1 enhancer null mice did not exhibit any of the characteristic morphological defects of the Pitx1−/− mutant. Our results indicate that Pitx1 expression is robust to the loss of its primary enhancer interaction, suggesting disruptions of regulatory topology at essential developmental genes may have mild phenotypic effects.