Regularized machine learning on molecular graph model explains systematic error in DFT enthalpies.
Ontology highlight
ABSTRACT:
SUBMITTER: Bhattacharjee H
PROVIDER: S-EPMC8277863 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:Enthalpies for bond-forming reactions that are subject to organocatalysis have been predicted using the high-accuracy CBS-QB3 model chemistry and six DFT functionals. Reaction enthalpies were decomposed into contributions from changes in bonding and other intramolecular effects via the hierarchy of homodesmotic reactions. The order of the reaction exothermicities (aldol < Mannich approximately alpha-aminoxylation) arises primarily from changes in formal bond types mediated by contributions from secondary intramolecular interactions. In each of these reaction types, methyl substitution at the beta- and gamma-positions stabilizes the products relative to the unsubstituted case. The performance of six DFT functionals (B3LYP, B3PW91, B1B95, MPW1PW91, PBE1PBE, and M06-2X), MP2, and SCS-MP2 has been assessed for the prediction of these reaction enthalpies. Even though the PBE1PBE and M06-2X functionals perform well for the aldol and Mannich reactions, errors roughly double when these functionals are applied to the alpha-aminoxylation reactions. B3PW91 and B1B95, which offer modest accuracy for the aldol and Mannich reactions, yield reliable predictions for the two alpha-aminoxylation reactions. The excellent performance of the M06-2X and PBE1PBE functionals for aldol and Mannich reactions stems from the cancellation of sizable errors arising from inadequate descriptions of the underlying bond transformations and intramolecular interactions. SCS-MP2/cc-pVTZ performs most consistently across these three classes of reactions, although the reaction exothermicities are systematically underestimated by 1-3 kcal mol(-1). Conventional MP2, when paired with the cc-pVTZ basis set, performs somewhat better than SCS-MP2 for some of these reactions, particularly the alpha-aminoxylations. Finally, the merits of benchmarking DFT functionals for the set of simple chemically meaningful transformations underlying all bond-forming reactions are discussed.
| S-EPMC2748113 | biostudies-literature
Project description:Accurate and automatic multi-needle detection in three-dimensional (3D) ultrasound (US) is a key step of treatment planning for US-guided brachytherapy. However, most current studies are concentrated on single-needle detection by only using a small number of images with a needle, regardless of the massive database of US images without needles. In this paper, we propose a workflow for multi-needle detection by considering the images without needles as auxiliary. Concretely, we train position-specific dictionaries on 3D overlapping patches of auxiliary images, where we develop an enhanced sparse dictionary learning method by integrating spatial continuity of 3D US, dubbed order-graph regularized dictionary learning. Using the learned dictionaries, target images are reconstructed to obtain residual pixels which are then clustered in every slice to yield centers. With the obtained centers, regions of interest (ROIs) are constructed via seeking cylinders. Finally, we detect needles by using the random sample consensus algorithm per ROI and then locate the tips by finding the sharp intensity drops along the detected axis for every needle. Extensive experiments were conducted on a phantom dataset and a prostate dataset of 70/21 patients without/with needles. Visualization and quantitative results show the effectiveness of our proposed workflow. Specifically, our method can correctly detect 95% of needles with a tip location error of 1.01 mm on the prostate dataset. This technique provides accurate multi-needle detection for US-guided HDR prostate brachytherapy, facilitating the clinical workflow.
| S-EPMC7370243 | biostudies-literature
Project description:Application of artificial intelligence (AI) in precision oncology typically involves predicting whether the cancer cells of a patient (previously unseen by AI models) will respond to any of a set of existing anticancer drugs, based on responses of previous training cell samples to those drugs. To expand the repertoire of anticancer drugs, AI has also been used to repurpose drugs that have not been tested in an anticancer setting, i.e., predicting the anticancer effects of a new drug on previously unseen cancer cells de novo. Here, we report a computational model that addresses both of the above tasks in a unified AI framework. Our model, referred to as deep learning-based graph regularized matrix factorization (DeepGRMF), integrates neural networks, graph models, and matrix-factorization techniques to utilize diverse information from drug chemical structures, their impact on cellular signaling systems, and cancer cell cellular states to predict cell response to drugs. DeepGRMF learns embeddings of drugs so that drugs sharing similar structures and mechanisms of action (MOAs) are closely related in the embedding space. Similarly, DeepGRMF also learns representation embeddings of cells such that cells sharing similar cellular states and drug responses are closely related. Evaluation of DeepGRMF and competing models on Genomics of Drug Sensitivity in Cancer (GDSC) and Cancer Cell Line Encyclopedia (CCLE) datasets show its superiority in prediction performance. Finally, we show that the model is capable of predicting effectiveness of a chemotherapy regimen on patient outcomes for the lung cancer patients in The Cancer Genome Atlas (TCGA) dataset*.
| S-EPMC8691529 | biostudies-literature
Project description:A micro-molecule of dimension 125 nm has caused around 479 million human infections (80 M for the USA) and 6.1 million human deaths (977,000 for the USA) worldwide and slashed the global economy by US$ 8.5 Trillion over two years period. The only other events in recent history that caused comparative human life loss through direct usage (either by human or nature, respectively) of structure-property relations of 'nano-structures' (either human-made or nature, respectively) were nuclear bomb attacks during World War II and 1918 Flu Pandemic. This molecule is called SARS-CoV-2, which causes a disease known as COVID-19. The high liability cost of the pandemic had incentivized various private, government, and academic entities to work towards finding a cure for this and emerging diseases. As an outcome, multiple vaccine candidates are discovered to avoid the infection in the first place. But so far, there has been no success in finding fully effective therapeutic candidates. In this paper, we attempted to provide multiple therapy candidates based upon a sophisticated multi-scale in-silico framework, which increases the probability of the candidates surviving an in-vivo trial. We have selected a group of ligands from the ZINC database based upon previously partially successful candidates, i.e., Hydroxychloroquine, Lopinavir, Remdesivir, Ritonavir. We have used the following robust framework to screen the ligands; Step-I: high throughput molecular docking, Step-II: molecular dynamics analysis, Step-III: density functional theory analysis. In total, we have analyzed 242,000(ligands)*9(proteins) = 2.178 million unique protein binding site/ligand combinations. The proteins were selected based on recent experimental studies evaluating potential inhibitor binding sites. Step-I had filtered that number down to 10 ligands/protein based on molecular docking binding energy, further screening down to 2 ligands/protein based on drug-likeness analysis. Additionally, these two ligands per protein were analyzed in Step-II with a molecular dynamic modeling-based RMSD filter of less than 1Å. It finally suggested three ligands (ZINC001176619532, ZINC000517580540, ZINC000952855827) attacking different binding sites of the same protein(7BV2), which were further analyzed in Step-III to find the rationale behind comparatively higher ligand efficacy.Supplementary informationThe online version contains supplementary material available at 10.1007/s10853-022-07195-8.
| S-EPMC9045684 | biostudies-literature
Project description:The application of machine learning to predict materials properties measured by experiments are valuable yet difficult due to the limited amount of experimental data. In this work, we use a multifidelity random forest model to learn the experimental formation enthalpy of materials with prediction accuracy higher than the Perdew-Burke-Ernzerhof (PBE) functional with linear correction, PBEsol, and meta-generalized gradient approximation (meta-GGA) functionals (SCAN and r2SCAN), and it outperforms the hotly studied deep neural network-based representation learning and transfer learning. We then use the model to calibrate the DFT formation enthalpy in the Materials Project database and discover materials with underestimated stability. The multifidelity model is also used as a data-mining approach to find how DFT deviates from experiments by explaining the model output.
| S-EPMC9516701 | biostudies-literature
Project description:Graph machine learning (GML) is receiving growing interest within the pharmaceutical and biotechnology industries for its ability to model biomolecular structures, the functional relationships between them, and integrate multi-omic datasets - amongst other data types. Herein, we present a multidisciplinary academic-industrial review of the topic within the context of drug discovery and development. After introducing key terms and modelling approaches, we move chronologically through the drug development pipeline to identify and summarize work incorporating: target identification, design of small molecules and biologics, and drug repurposing. Whilst the field is still emerging, key milestones including repurposed drugs entering in vivo studies, suggest GML will become a modelling framework of choice within biomedical machine learning.
| S-EPMC8574649 | biostudies-literature
Project description:The efficient calculation of nucleation collective variables (CVs) is one of the main limitations to the application of enhanced sampling methods to the investigation of nucleation processes in realistic environments. Here we discuss the development of a graph-based model for the approximation of nucleation CVs that enables orders-of-magnitude gains in computational efficiency in the on-the-fly evaluation of nucleation CVs. By performing simulations on a nucleating colloidal system mimicking a multistep nucleation process from solution, we assess the model's efficiency in both postprocessing and on-the-fly biasing of nucleation trajectories with pulling, umbrella sampling, and metadynamics simulations. Moreover, we probe and discuss the transferability of graph-based models of nucleation CVs across systems using the model of a CV based on sixth-order Steinhardt parameters trained on a colloidal system to drive the nucleation of crystalline copper from its melt. Our approach is general and potentially transferable to more complex systems as well as to different CVs.
| S-EPMC10902841 | biostudies-literature
Project description:The human brain has been an object of extensive investigation in different fields. While several studies have focused on understanding the neural correlates of error processing, advances in brain-machine interface systems using non-invasive techniques further enabled the use of the measured signals in different applications. The possibility of detecting these error-related potentials (ErrPs) under different experimental setups on a single-trial basis has further increased interest in their integration in closed-loop settings to improve system performance, for example, by performing error correction. Fewer works have, however, aimed at reducing future mistakes or learning. We present a review focused on the current literature using non-invasive systems that have combined the ErrPs information specifically in a reinforcement learning framework to go beyond error correction and have used these signals for learning.
| S-EPMC9263570 | biostudies-literature
Project description:Machine learning continues to make strident advances in the prediction of desired properties concerning drug development. Problematically, the efficacy of machine learning in these arenas is reliant upon highly accurate and abundant data. These two limitations, high accuracy and abundance, are often taken together; however, insight into the dataset accuracy limitation of contemporary machine learning algorithms may yield insight into whether non-bench experimental sources of data may be used to generate useful machine learning models where there is a paucity of experimental data. We took highly accurate data across six kinase types, one GPCR, one polymerase, a human protease, and HIV protease, and intentionally introduced error at varying population proportions in the datasets for each target. With the generated error in the data, we explored how the retrospective accuracy of a Naïve Bayes Network, a Random Forest Model, and a Probabilistic Neural Network model decayed as a function of error. Additionally, we explored the ability of a training dataset with an error profile resembling that produced by the Free Energy Perturbation method (FEP+) to generate machine learning models with useful retrospective capabilities. The categorical error tolerance was quite high for a Naïve Bayes Network algorithm averaging 39% error in the training set required to lose predictivity on the test set. Additionally, a Random Forest tolerated a significant degree of categorical error introduced into the training set with an average error of 29% required to lose predictivity. However, we found the Probabilistic Neural Network algorithm did not tolerate as much categorical error requiring an average of 20% error to lose predictivity. Finally, we found that a Naïve Bayes Network and a Random Forest could both use datasets with an error profile resembling that of FEP+. This work demonstrates that computational methods of known error distribution like FEP+ may be useful in generating machine learning models not based on extensive and expensive in vitro-generated datasets.
| S-EPMC6601015 | biostudies-literature
Project description:Antioxidant peptides (AOPs) are highly valued in food and pharmaceutical industries due to their significant role in human function. This study introduces a novel approach to identifying robust AOPs using a deep generative model based on sequence representation. Through filtration with a deep-learning classification model and subsequent clustering via the Butina cluster algorithm, twelve peptides (GP1-GP12) with potential antioxidant capacity were predicted. Density functional theory (DFT) calculations guided the selection of six peptides for synthesis and biological experiments. Molecular orbital representations revealed that the HOMO for these peptides is primarily localized on the indole segment, underscoring its pivotal role in antioxidant activity. All six synthesized peptides exhibited antioxidant activity in the DPPH assay, while the hydroxyl radical test showed suboptimal results. A hemolysis assay confirmed the non-hemolytic nature of the generated peptides. Additionally, an in silico investigation explored the potential inhibitory interaction between the peptides and the Keap1 protein. Analysis revealed that ligands GP3, GP4, and GP12 induced significant structural changes in proteins, affecting their stability and flexibility. These findings highlight the capability of machine learning approaches in generating novel antioxidant peptides.
| S-EPMC10948870 | biostudies-literature