A5046241574- Computational Drug Discovery Methods
- Protein Structure and Dynamics
- Machine Learning in Materials Science
- Gene expression and cancer classification
- vaccines and immunoinformatics approaches
- Lipid Membrane Structure and Behavior
- Genetics, Bioinformatics, and Biomedical Research
- Cell Image Analysis Techniques
- Computer Graphics and Visualization Techniques
- Metal-Organic Frameworks: Synthesis and Applications
- Medical Imaging Techniques and Applications
- 3D Surveying and Cultural Heritage
- Spectroscopy and Quantum Chemical Studies
- Covalent Organic Framework Applications
- Remote Sensing and LiDAR Applications
- SARS-CoV-2 and COVID-19 Research
- Advanced MRI Techniques and Applications
- Force Microscopy Techniques and Applications
- Machine Learning in Bioinformatics
- Analytical Chemistry and Chromatography
- Electrochemical Analysis and Applications
- Biosensors and Analytical Detection
- Advanced biosensing and bioanalysis techniques
- Advanced X-ray and CT Imaging
- Simulation Techniques and Applications
Lawrence Livermore National Security
2020-2025
Lawrence Livermore National Laboratory
2018-2025
Abstract To enable personalized cancer treatment, machine learning models have been developed to predict drug response as a function of tumor and features. However, most algorithm development efforts relied on cross-validation within single study assess model accuracy. While an essential first step, biological data set typically provides overly optimistic estimate the prediction performance independent test sets. provide more rigorous assessment generalizability between different studies, we...
Limited-Angle Computed Tomography (LACT) is a nondestructive 3D imaging technique used in variety of applications ranging from security to medicine. The limited angle coverage LACT often dominant source severe artifacts the reconstructed images, making it challenging inverse problem. Diffusion models are recent class deep generative for synthesizing realistic images using image denoisers. In this work, we present DOLCE as first framework integrating conditionally-trained diffusion and...
Identifying conductive metal–organic frameworks (MOFs) with a coupled ion-electron behavior from vast array of existing MOFs offers cost-effective strategy to tap into their potential in energy storage applications. This study employs classification and regression machine learning (ML) rapidly screen the CoREMOF database experimental methodologies validate ML predictions. process revealed structure–property relationships contributing MOFs' bulk conductivity. Among 60 compounds predicted,...
Accurately predicting small molecule partitioning and hydrophobicity is critical in the drug discovery process. There are many heterogeneous chemical environments within a cell entire human body. For example, drugs must be able to cross hydrophobic cellular membrane reach their intracellular targets, an important driving force for drug–protein binding. Atomistic molecular dynamics (MD) simulations routinely used calculate free energies of molecules binding proteins, crossing lipid membranes,...
Recent advances in 3D structure-based deep learning approaches demonstrate improved accuracy predicting protein-ligand binding affinity drug discovery. These methods complement physics-based computational modeling such as molecular docking for virtual high-throughput screening. Despite recent and predictive performance, most this category primarily rely on utilizing co-crystal complex structures experimentally measured affinities both input output data model training. Nevertheless, are not...
A rapid response is necessary to contain emergent biological outbreaks before they can become pandemics. The novel coronavirus (SARS-CoV-2) that causes COVID-19 was first reported in December of 2019 Wuhan, China and reached most corners the globe less than two months. In just over a year since initial infections, infected almost 100 million people worldwide. Although similar SARS-CoV MERS-CoV, SARS-CoV-2 has resisted treatments are effective against other coronaviruses. Crystal structures...
We investigated gramicidin A (gA) subunit dimerization in lipid bilayers using microsecond-long replica-exchange umbrella sampling simulations, millisecond-long unbiased molecular dynamics and machine learning. Our simulations led to a dimer structure that is indistinguishable from the experimentally determined gA channel structures, with two subunits joined by six hydrogen bonds (6HB). The also uncovered additional different gA–gA stacking orientations were stabilized four or (4HB 2HB)....
Structure-based Deep Fusion models were recently shown to outperform several physics- and machine learning-based protein-ligand binding affinity prediction methods. As part of a multi-institutional COVID-19 pandemic response, over 500 million small molecules computationally screened against four protein structures from the novel coronavirus (SARS-CoV-2), which causes COVID-19. Three enhancements made in order evaluate more than 5 billion docked poses on SARS-CoV-2 targets. First, concept was...
Partitioning of bioactive molecules, including drugs, into cell membranes may produce indiscriminate changes in membrane protein function. As a guide to safe drug development, it therefore becomes important be able predict the bilayer-perturbing potency hydrophobic/amphiphilic drugs candidates. Toward this end, we exploited gramicidin channels as molecular force probes and developed silico vitro assays measure drugs' bilayer-modifying potency. We examined eight drug-like molecules that were...
Automated 3D modeling of building interiors is used in applications such as virtual reality and environment mapping. Texturing these models allows for photo-realistic visualizations the data collected by systems. While acquisition times mobile mapping systems are considerably shorter than static ones, their recovered camera poses often suffer from inaccuracies, resulting visible discontinuities when successive images projected onto a surface texturing. We present method texture indoor...
Author(s): Simons, Lance C.; He, Stewart; Tittmann, Peter; Amenta, Nina | Abstract: Airborne Light Detection And Ranging (LiDAR) is an increasingly important modality for remote sensing of forests. Unfortunately, the lack smooth surfaces complicates visualization LiDAR data and results fundamental analysis tasks that interest environmental scientists. In this paper, we use multi-pass point-cloud rendering to produce shadows, approximate occlusion, a non-photorealistic silhouette effect which...
Metal-organic frameworks (MOFs) are highly versatile structures composed of metal ions linked by organic molecules. They offer a broad range applications in electronics and light-based technologies due to their flexibility. However, the connection between components often hinders efficient charge movement, which is essential for electrical conduction. To address this challenge, researchers have been innovatively designing MOFs enhance conductivity, typically low. This process complex...
To enable personalized cancer treatment, machine learning models have been developed to predict drug response as a function of tumor and features. However, most algorithm development efforts relied on cross validation within single study assess model accuracy. While an essential first step, biological data set typically provides overly optimistic estimate the prediction performance independent test sets. provide more rigorous assessment generalizability between different studies, we use...
Limited-Angle Computed Tomography (LACT) is a non-destructive evaluation technique used in variety of applications ranging from security to medicine. The limited angle coverage LACT often dominant source severe artifacts the reconstructed images, making it challenging inverse problem. We present DOLCE, new deep model-based framework for that uses conditional diffusion model as an image prior. Diffusion models are recent class generative relatively easy train due their implementation...
Structure-based Deep Fusion models were recently shown to outperform several physics- and machine learning-based protein-ligand binding affinity prediction methods. As part of a multi-institutional COVID-19 pandemic response, over 500 million small molecules computationally screened against four protein structures from the novel coronavirus (SARS-CoV-2), which causes COVID-19. Three enhancements made in order evaluate more than 5 billion docked poses on SARS-CoV-2 targets. First, concept was...
Abstract Background Predicting molecular activity against protein targets is difficult because of the paucity experimental data. Approaches like multitask modeling and collaborative filtering seek to improve model accuracy by leveraging results from multiple targets, but are limited different compounds measured with assays, leading sparse data matrices. Profile-QSAR (pQSAR) 2.0 addresses this problem fitting a series partial least squares models for each target, using as features predictions...