Predicting Structural Properties of Pure Silica Zeolites Using Deep Neural Network Potentials
Ontology highlight
ABSTRACT:
SUBMITTER:
PROVIDER: S-EPMC9885523 | biostudies-literature | 2023 Jan
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Several computational techniques for solid-state applications have recently been proposed to enlarge the scope of computer simulations of large molecular systems. In this contribution, we focused on two of these, namely, HF-3c and PBEh-3c. They were recently proposed by the Grimme's group, as "low-cost" ab initio-based techniques for electronic structure calculation of large systems and were proved to be effective essentially for organic molecules. HF-3c is based on a Hartree-Fock Hamiltonian with a minimal Gaussian quality basis set, whereas PBEh-3c is a density functional theory (DFT) based method with a hybrid functional and a medium-quality basis set. Both HF-3c and PBEh-3c account for dispersion (London) interactions and are free from the basis set superposition error due to limited basis set size, through several pairwise semiempirical corrections. To the best of our knowledge, despite the promising results on the cost-accuracy side of molecular simulations of organic molecules, these methods have been used only in few cases for solid-state applications. In this contribution, we studied the performance of HF-3c and PBEh-3c for predicting the properties of inorganic crystals to enlarge the applicability of these cheap and fast methodologies. As a testing ground, we have chosen a well-known class of material, e.g., microporous all-silica zeolites. We benchmarked geometries, formation energies, vibrational features, and mechanical properties by comparing the results with literature data from both experiment and computer simulation. For structures, HF-3c is extremely accurate in predicting the zeolites cell volume, albeit we do not include any vibrational contribution, neither zero point nor thermal, on the computed volumes, which may introduce small variations in the predicted values. For the energetic, the relative stability of the zeolites using the DFT//HF-3c approach allows predictions within the experimental error for most of the cases taken into consideration when the experimental enthalpies were corrected back to electronic energies by using the HF-3c thermodynamic contributions computed in the harmonic approximation. This strategy is particularly convenient, as the slow step (geometry optimization) is carried out with the cheapest HF-3c method, whereas the fast step (single point energy evaluation) is carried out with costly DFT methods. In this sense, the use of the DFT//HF-3c approach results to be a promising one to predict the stability and structure of microporous materials. Finally, the HF-3c method predicts the mechanical properties of the zeolite set in reasonable agreement with respect to those computed with the state-of-the-art DFT simulations, indicating the HF-3c method as a possible technique for the mechanical stability screenings of microporous materials.
| S-EPMC6648702 | biostudies-literature
Project description:Drug-target interactions (DTIs) prediction is an important step in drug discovery. As traditional biological experiments or high-throughput screening are high cost and time-consuming, many deep learning models have been developed. Overfitting must be avoided when training deep learning models. We propose a simple framework, called OverfitDTI, for DTI prediction. In OverfitDTI, a deep neural network (DNN) model is overfit to sufficiently learn the features of the chemical space of drugs and the biological space of targets. The weights of trained DNN model form an implicit representation of the nonlinear relationship between drugs and targets. Performance of OverfitDTI on three public datasets showed that the overfit DNN models fit the nonlinear relationship with high accuracy. We identified fifteen compounds that interacted with TEK, a receptor tyrosine kinase contributing to vascular homeostasis, and the predicted AT9283 and dorsomorphin were experimentally demonstrated as inhibitors of TEK in human umbilical vein endothelial cells (HUVECs).
| S-EPMC10480310 | biostudies-literature
Project description:Alternative polyadenylation (APA) is a major driver of transcriptome diversity in human cells. Here, we use deep learning to predict APA from DNA sequence alone. We trained our model (APARENT, APA REgression NeT) on isoform expression data from over 3 million APA reporters. APARENT's predictions are highly accurate when tasked with inferring APA in synthetic and human 3'UTRs. Visualizing features learned across all network layers reveals that APARENT recognizes sequence motifs known to recruit APA regulators, discovers previously unknown sequence determinants of 3' end processing, and integrates these features into a comprehensive, interpretable, cis-regulatory code. We apply APARENT to forward engineer functional polyadenylation signals with precisely defined cleavage position and isoform usage and validate predictions experimentally. Finally, we use APARENT to quantify the impact of genetic variants on APA. Our approach detects pathogenic variants in a wide range of disease contexts, expanding our understanding of the genetic origins of disease.
| S-EPMC6599575 | biostudies-literature
Project description:BackgroundEssential genes are those genes that are critical for the survival of an organism. The prediction of essential genes in bacteria can provide targets for the design of novel antibiotic compounds or antimicrobial strategies.ResultsWe propose a deep neural network for predicting essential genes in microbes. Our architecture called DEEPLYESSENTIAL makes minimal assumptions about the input data (i.e., it only uses gene primary sequence and the corresponding protein sequence) to carry out the prediction thus maximizing its practical application compared to existing predictors that require structural or topological features which might not be readily available. We also expose and study a hidden performance bias that effected previous classifiers. Extensive results show that DEEPLYESSENTIAL outperform existing classifiers that either employ down-sampling to balance the training set or use clustering to exclude multiple copies of orthologous genes.ConclusionDeep neural network architectures can efficiently predict whether a microbial gene is essential (or not) using only its sequence information.
| S-EPMC7525945 | biostudies-literature
Project description:Accurate protein structure prediction from amino acid sequence is still an unsolved problem. The most reliable methods centre on template based modelling. However, the accuracy of these models entirely depends on the availability of experimentally resolved homologous template structures. In order to generate more accurate models, extensive physics based molecular dynamics (MD) refinement simulations are performed to sample many different conformations to find improved conformational states. In this study, we propose a deep recurrent network model, called DeepTrajectory, that is able to identify these improved conformational states, with high precision, from a variety of different MD based sampling protocols. The proposed model learns the temporal patterns of features computed from MD trajectory data in order to classify whether each recorded simulation snapshot is an improved quality conformational state, decreased quality conformational state or whether there is no perceivable change in state with respect to the starting conformation. The model was trained and tested on 904 trajectories from 42 different protein systems with a cumulative number of more than 1.7 million snapshots. We show that our model outperforms other state of the art machine-learning algorithms that do not consider temporal dependencies. To our knowledge, DeepTrajectory is the first implementation of a time-dependent deep-learning protocol that is re-trainable and able to adapt to any new MD based sampling procedure, thereby demonstrating how a neural network can be used to learn the latter part of the protein folding funnel.
| S-EPMC6122789 | biostudies-literature
Project description:We have developed a new deep boosted molecular dynamics (DBMD) method. Probabilistic Bayesian neural network models were implemented to construct boost potentials that exhibit Gaussian distribution with minimized anharmonicity, thereby allowing for accurate energetic reweighting and enhanced sampling of molecular simulations. DBMD was demonstrated on model systems of alanine dipeptide and the fast-folding protein and RNA structures. For alanine dipeptide, 30 ns DBMD simulations captured up to 83-125 times more backbone dihedral transitions than 1 μs conventional molecular dynamics (cMD) simulations and were able to accurately reproduce the original free energy profiles. Moreover, DBMD sampled multiple folding and unfolding events within 300 ns simulations of the chignolin model protein and identified low-energy conformational states comparable to previous simulation findings. Finally, DBMD captured a general folding pathway of three hairpin RNAs with the GCAA, GAAA, and UUCG tetraloops. Based on a deep learning neural network, DBMD provides a powerful and generally applicable approach to boosting biomolecular simulations. DBMD is available with open source in OpenMM at https://github.com/MiaoLab20/DBMD/.
| S-EPMC10355842 | biostudies-literature
Project description:Untargeted metabolomic measurements using mass spectrometry are a powerful tool for uncovering new small molecules with environmental and biological importance. The small molecule identification step, however, still remains an enormous challenge due to fragmentation difficulties or unspecific fragment ion information. Current methods to address this challenge are often dependent on databases or require the use of nuclear magnetic resonance (NMR), which have their own difficulties. The use of the gas-phase collision cross section (CCS) values obtained from ion mobility spectrometry (IMS) measurements were recently demonstrated to reduce the number of false positive metabolite identifications. While promising, the amount of empirical CCS information currently available is limited, thus predictive CCS methods need to be developed. In this article, we expand upon current experimental IMS capabilities by predicting the CCS values using a deep learning algorithm. We successfully developed and trained a prediction model for CCS values requiring only information about a compound's SMILES notation and ion type. The use of data from five different laboratories using different instruments allowed the algorithm to be trained and tested on more than 2400 molecules. The resulting CCS predictions were found to achieve a coefficient of determination of 0.97 and median relative error of 2.7% for a wide range of molecules. Furthermore, the method requires only a small amount of processing power to predict CCS values. Considering the performance, time, and resources necessary, as well as its applicability to a variety of molecules, this model was able to outperform all currently available CCS prediction algorithms.
| S-EPMC6628689 | biostudies-literature
Project description:Cancer origin determination combined with site-specific treatment of metastatic cancer patients is critical to improve patient outcomes. Existing pathology and gene expression-based techniques often have limited performance. In this study, we developed a deep neural network (DNN)-based classifier for cancer origin prediction using DNA methylation data of 7,339 patients of 18 different cancer origins from The Cancer Genome Atlas (TCGA). This DNN model was evaluated using four strategies: (1) when evaluated by 10-fold cross-validation, it achieved an overall specificity of 99.72% (95% CI 99.69%-99.75%) and sensitivity of 92.59% (95% CI 91.87%-93.30%); (2) when tested on hold-out testing data of 1,468 patients, the model had an overall specificity of 99.83% and sensitivity of 95.95%; (3) when tested on 143 metastasized cancer patients (12 cancer origins), the model achieved an overall specificity of 99.47% and sensitivity of 95.95%; and (4) when tested on an independent dataset of 581 samples (10 cancer origins), the model achieved overall specificity of 99.91% and sensitivity of 93.43%. Compared to existing pathology and gene expression-based techniques, the DNA methylation-based DNN classifier showed higher performance and had the unique advantage of easy implementation in clinical settings. In summary, our study shows that DNA methylation-based DNN models has potential in both diagnosis of cancer of unknown primary and identification of cancer cell types of circulating tumor cells.
| S-EPMC7209244 | biostudies-literature
Project description:BackgroundThe availability of high-throughput omics datasets from large patient cohorts has allowed the development of methods that aim at predicting patient clinical outcomes, such as survival and disease recurrence. Such methods are also important to better understand the biological mechanisms underlying disease etiology and development, as well as treatment responses. Recently, different predictive models, relying on distinct algorithms (including Support Vector Machines and Random Forests) have been investigated. In this context, deep learning strategies are of special interest due to their demonstrated superior performance over a wide range of problems and datasets. One of the main challenges of such strategies is the "small n large p" problem. Indeed, omics datasets typically consist of small numbers of samples and large numbers of features relative to typical deep learning datasets. Neural networks usually tackle this problem through feature selection or by including additional constraints during the learning process.MethodsWe propose to tackle this problem with a novel strategy that relies on a graph-based method for feature extraction, coupled with a deep neural network for clinical outcome prediction. The omics data are first represented as graphs whose nodes represent patients, and edges represent correlations between the patients' omics profiles. Topological features, such as centralities, are then extracted from these graphs for every node. Lastly, these features are used as input to train and test various classifiers.ResultsWe apply this strategy to four neuroblastoma datasets and observe that models based on neural networks are more accurate than state of the art models (DNN: 85%-87%, SVM/RF: 75%-82%). We explore how different parameters and configurations are selected in order to overcome the effects of the small data problem as well as the curse of dimensionality.ConclusionsOur results indicate that the deep neural networks capture complex features in the data that help predicting patient clinical outcomes.
| S-EPMC6923884 | biostudies-literature
Project description:We have developed a new Deep Boosted Molecular Dynamics (DBMD) method. Probabilistic Bayesian neural network models were implemented to construct boost potentials that exhibit Gaussian distribution with minimized anharmonicity, thereby allowing for accurate energetic reweighting and enhanced sampling of molecular simulations. DBMD was demonstrated on model systems of alanine dipeptide and the fast-folding protein and RNA structures. For alanine dipeptide, 30ns DBMD simulations captured up to 83-125 times more backbone dihedral transitions than 1μs conventional molecular dynamics (cMD) simulations and were able to accurately reproduce the original free energy profiles. Moreover, DBMD sampled multiple folding and unfolding events within 300ns simulations of the chignolin model protein and identified low-energy conformational states comparable to previous simulation findings. Finally, DBMD captured a general folding pathway of three hairpin RNAs with the GCAA, GAAA, and UUCG tetraloops. Based on Deep Learning neural network, DBMD provides a powerful and generally applicable approach to boosting biomolecular simulations. DBMD is available with open source in OpenMM at https://github.com/MiaoLab20/DBMD/.
| S-EPMC10081221 | biostudies-literature