A5035522231- Machine Learning in Materials Science
- Advanced Chemical Physics Studies
- Computational Drug Discovery Methods
- Protein Structure and Dynamics
- Cancer-related molecular mechanisms research
- Osteoarthritis Treatment and Mechanisms
- Genetic Associations and Epidemiology
- Genomics and Chromatin Dynamics
- MicroRNA in disease regulation
- Bioinformatics and Genomic Networks
- Epigenetics and DNA Methylation
- Psychosomatic Disorders and Their Treatments
- Electronic and Structural Properties of Oxides
- Phytochemicals and Antioxidant Activities
- Aerogels and thermal insulation
- Pharmacological Effects of Natural Compounds
- Bone Metabolism and Diseases
- Glass properties and applications
- Topic Modeling
- X-ray Diffraction in Crystallography
- Invertebrate Immune Response Mechanisms
- Natural Antidiabetic Agents Studies
- Inflammatory mediators and NSAID effects
- Quantum and electron transport phenomena
- Circular RNAs in diseases
Princeton University
2019-2025
Shandong First Medical University
2025
Universidad del Noreste
2025
Peking University Shenzhen Hospital
2022-2024
Flatiron Health (United States)
2024
Flatiron Institute
2024
Inner Mongolia Agricultural University
2024
Wuhan Polytechnic University
2024
Donghua University
2024
Chengdu University of Traditional Chinese Medicine
2023-2024
DeePMD-kit is a powerful open-source software package that facilitates molecular dynamics simulations using machine learning potentials known as Deep Potential (DP) models. This package, which was released in 2017, has been widely used the fields of physics, chemistry, biology, and material science for studying atomistic systems. The current version offers numerous advanced features, such DeepPot-SE, attention-based hybrid descriptors, ability to fit tensile properties, type embedding, model...
Developing an efficient method for determination of the DNA-binding proteins, due to their vital roles in gene regulation, is becoming highly desired since it would be invaluable advance our understanding protein functions. In this study, we proposed a new prediction by performing feature rank using random forest and wrapper-based selection forward best-first search strategy. The features comprise information from primary sequence, predicted secondary structure, relative solvent...
Abstract There is a debate on whether invertebrates possess an antiviral immunity similar to the interferon (IFN) system of vertebrates. The Vago gene from arthropods encodes viral-activated secreted peptide that restricts virus infection through activating JAK-STAT pathway and considered be cytokine functionally IFN. In this study, first crustacean IFN regulatory factor (IRF)-like was identified in Pacific white shrimp, Litopenaeus vannamei . L. IRF showed protein nature mammalian IRFs...
Machine-learning-based interatomic potential energy surface (PES) models are revolutionizing the field of molecular modeling. However, although much faster than electronic structure schemes, these suffer from costly computations via deep neural networks to predict and atomic forces, resulting in lower running efficiency as compared typical empirical force fields. Herein, we report a model compression scheme for boosting performance Deep Potential (DP) model, learning-based PES model. This...
The plastic deformation of crystalline materials can be understood by considering their structural defects such as disclinations and dislocations. Although also glasses are solids, structure resembles closely the one a liquid hence concept becomes ill-defined. As consequence it is very challenging to rationalize on microscopic level mechanical properties close yielding point relate events properties. Here we investigate topological characteristics eigenvector field vibrational excitations...
Incorporating passive heating structures into personal thermal management technologies could effectively mitigate the escalating energy crisis. However, current materials struggle to balance thickness and insulating capability, resulting in compromised comfort, space efficiency, limited thermoregulatory performance. Here, a dual air-gelation strategy, is developed directly synthesize ultrathin self-sustainable metafabric with 3D dual-network structure during electrospinning. Controlling...
We propose a general machine learning-based framework for building an accurate and widely-applicable energy functional within the of generalized Kohn-Sham density theory. To this end, we develop way training self-consistent models that are capable taking large datasets from different systems kinds labels. demonstrate results procedure gives chemically predictions on energy, force, dipole, electron class molecules. It can be continuously improved when more data available.
We introduce the deep post Hartree-Fock (DeePHF) method, a machine learning-based scheme for constructing accurate and transferable models ground-state energy of electronic structure problems. DeePHF predicts difference between results highly such as coupled cluster method low accuracy (HF) using orbitals input. It preserves all symmetries original high model. The added computational cost is less than that reference HF or DFT scales linearly with respect to system size. examine performance...
The nuclear factor-kappa B (NF-κB) pathways play important roles in innate immune responses. IκB is the main cytoplasmic inhibitor of NF-κB. In this study, we identified LvCactus gene from Litopenaeus vannamei, which first cloned homologue subphylum Crustacea. contains six predicted ankyrin repeats, show similarities to those Cactus proteins insects. localizes cytoplasm and interacts with LvDorsal, an L. vannamei Drosophila melanogaster Dorsal belonging class II NF-κB family, prevent its...
This study demonstrates the effect and DNA methylation-related mechanisms of a high-salt diet salt memory-induced hypertension vasculopathy. Thirty Sprague Dawley rats were randomly divided into control (CON) group (n = 6) modeling 24). A 12% NaCl solution (1 mL/100 g) was intragastrically administered for 60 consecutive days modeling. An increase in blood pressure up to 140 mmHg considered successful Twelve fifteen successfully modeled selected High Salt Diet (HSD) Memory (HSM) 6). Rats HSD...
This article investigates the critical issue of dataset bias in medical imaging, with a particular emphasis on racial disparities caused by uneven population distribution collection. Our analysis reveals that segmentation datasets are significantly biased, primarily influenced demographic composition their collection sites. For instance, Scanning Laser Ophthalmoscopy (SLO) fundus collected United States predominantly feature images White individuals, minority groups underrepresented....
Large language models have achieved remarkable success in various tasks but suffer from high computational costs during inference, limiting their deployment resource-constrained applications. To address this issue, we propose a novel CITER (\textbf{C}ollaborative \textbf{I}nference with \textbf{T}oken-l\textbf{E}vel \textbf{R}outing) framework that enables efficient collaboration between small and large (SLMs & LLMs) through token-level routing strategy. Specifically, routes non-critical...
Adversarial attacks are widely used to evaluate model robustness, and in black-box scenarios, the transferability of these becomes crucial. Existing generator-based have excellent generalization due their instance-agnostic nature. However, when training generators for multi-target tasks, success rate transfer is relatively low limitations model's capacity. To address challenges, we propose a novel Dual-Flow framework adversarial attacks, utilizing Cascading Distribution Shift Training...
This study introduces MedVis Suite, a framework developed to address key challenges in medical image analysis using MRI scans. Suite integrates advanced machine learning techniques, including U-Net-based segmentation model optimized for bone segmentation, and 3D reconstruction capabilities. An in-depth evaluation of is performed across anatomical planes, optimizing both loss functions scales. The axial view showed the highest performance with Dice score 0.91 baseline model, while combination...
We report a molecular dynamics study of quality the ferroelectric phase transition in crystalline <a:math xmlns:a="http://www.w3.org/1998/Math/MathML"><a:mrow><a:mi>PbTi</a:mi><a:msub><a:mtext>O</a:mtext><a:mn>3</a:mn></a:msub></a:mrow></a:math>. model anharmonicity accurately terms potential energy and polarization surfaces trained on density functional theory data with modern machine learning techniques. Our simulations demonstrate that has strong order-disorder character, agreement...
High-performance computing, together with a neural network model trained from data generated first-principles methods, has greatly boosted applications of ab initio molecular dynamics in terms spatial and temporal scales on modern supercomputers. Previous state-of-the-art can achieve 1 -- 2 nanoseconds simulation per day for 100-million atoms the entire Summit supercomputer. In this paper, we have significantly reduced memory footprint computational time by comprehensive approach both...
Nigella sativa is a valuable herb for its functional compositions in both food and medication. N. seeds can enhance immunity, anti-inflammation analgesia hypoglycemia, but most of the related researches are to volatile oil extracts, activity mechanism compounds not clear. In this study, Ethyl-α-D-galactopyranoside (EG), Methyl-α-D-glucoside (MG), 3-O-[β-D-xylopyranose-(1 → 3)-α-L-rhamnose-(1 2)-α-L-arabinose]-28-O-[α-L-rhamnose-(1 4)-β-D-glucopyranose-L-(1 6)-β-D-glucopyranose]-hederagenin...