Results 101 to 110 of about 38,896 (265)

Application of Graph Spatio-Temporal Convolutional Neural Network in Motor Intention Recognition of Stroke Patients

康复学报
ObjectiveTo evaluate the decoding accuracy and model performance of a graph spatio-temporal convolutional neural network (G-STCNN) in motor intention recognition of stroke patients.MethodsWe developed a novel G-STCNN model by integrating graph ...
XU Hui, MA Jun, ZHANG Jingjing, ZHANG Tao, ZHAO Zhiwei, SHAN Chunlei   +5 more
doaj  

STAID: A Self‐Refining Deep Learning Framework for Spatial Cell‐Type Deconvolution with Biologically Informed Modeling

Advanced Science, EarlyView.
STAID is a unified deep learning framework that couples iterative pseudo‐spot refinement with neural network training through a feedback loop and exploits gene co‐expression information to model higher‐order interactions, achieving accurate and robust cell‐type deconvolution in spatial transcriptomics.
Jixin Liu, Shuli Sun, Zhengliang Lv, Xinyu Liu, Yihua Wang, Bingqiang Liu   +5 more
wiley   +1 more source

Diffusion‐Based Generative Model With Scaffold‐Hopping Strategy Yields Highly Potent Bioactive Molecules

Advanced Science, EarlyView.
SMarT‐Diff introduces a multi‐objective generative paradigm that integrates scaffold hopping with structure‐aware scoring to enable controlled exploration beyond the training distribution. The framework consistently balances drug‐likeness, synthesizes accessibility and bioactivity, yielding chemically diverse candidates with enhanced properties.
Yuwei Yang, Xiaoqing Gong, Shukai Gu, Jing Li, Bo Liu, Yanan Tian, Qianqian Zhang, Xiaojun Yao, Huanxiang Liu   +8 more
wiley   +1 more source

Topology‐Aware Deep Learning on Higher‐Order Structures for Drug Response Prediction

Advanced Science, EarlyView.
We present TopDr, a topology‐aware deep learning framework that encodes both drugs and cell lines as multiscale simplicial complexes, capturing interactions at the 0‐, 1‐, and 2‐simplex levels. By jointly integrating local higher‐order neighborhoods and global topological structures, TopDr generates enriched representations for sensitivity prediction ...
Cong Shen, Guancen Lin, Chuan‐Shen Hu, Yong Wang   +3 more
wiley   +1 more source

De Novo Design of Membrane‐Targeting Antimicrobial Peptides Against Gram‐Negative Bacteria Using a Generative Artificial Intelligence Framework

Advanced Science, EarlyView.
Antimicrobial resistance caused by Gram‐negative bacteria remains difficult to overcome due to the protective outer membrane. To address this challenge, a multi‐condition constrained generative AI framework, GenMTAMP is proposed for de novo membrane‐targeting antimicrobial peptide design by integrating physicochemical and spatial structure descriptors.
Jingxiao Yu, Da‐Wen Sun, QingYi Wei, Yibing Zhang, Jidong Tang, Hongbin Pu   +5 more
wiley   +1 more source

ProSiteHunter: A Unified Framework for Sequence‐Based Prediction of Protein‐Nucleic Acid and Protein‐Protein Binding Sites

Advanced Science, EarlyView.
This study proposed a unified sequence‐based framework for protein binding site prediction, which adopted a tri‐track semantic multi‐source feature fusion strategy to effectively capture diverse macromolecular interaction sites and further improved the accuracy of antibody‐antigen interaction prediction.
Dongliang Hou, Qihang Zhen, Zexin Lv, Xinyue Cui, Suhui Wang, Minghua Hou, Zhan Zhou, Xiaogen Zhou, Guijun Zhang   +8 more
wiley   +1 more source

Construction of Sabatier Volcanoes for CO2 Hydrogenation to C1‐2 Oxygenates Using Data‐Efficient Machine Learning

Advanced Science, EarlyView.
A new data‐efficient framework combining DFT calculations, a neural network model, and automated graph analysis of catalytic reaction networks is proposed and applied to CO2 hydrogenation on transition metal nanoparticles. The analysis shows how efficient C2 oxygenate production requires a balance between CHx formation, C–C coupling, protonation, and ...
Mikhail V. Polynski, Sergey M. Kozlov
wiley   +1 more source

Non-convolutional graph neural networks.

Advances in Neural Information Processing Systems 37
Rethink convolution-based graph neural networks (GNN) -- they characteristically suffer from limited expressiveness, over-smoothing, and over-squashing, and require specialized sparse kernels for efficient computation. Here, we design a simple graph learning module entirely free of convolution operators, coined random walk with unifying memory (RUM ...
Yuanqing Wang, Kyunghyun Cho
openaire   +3 more sources

Fusing graph transformer with multi-aggregate GCN for enhanced drug–disease associations prediction

BMC Bioinformatics
Background Identification of potential drug–disease associations is important for both the discovery of new indications for drugs and for the reduction of unknown adverse drug reactions. Exploring the potential links between drugs and diseases is crucial
Shihui He, Lijun Yun, Haicheng Yi
doaj   +1 more source

Accurately Deciphering Tissue Heterogeneity From Spatial Multi‐Modal and Multi‐Omics With STransformer

Advanced Science, EarlyView.
STransformer is a unified deep learning framework designed to seamlessly accommodate a comprehensive landscape of spatial data. By simultaneously capturing short‐range cellular interactions and tissue‐wide semantic patterns, it extracts robust representations to accurately dissect complex tissue heterogeneity.
Xingyi Li, Jialuo Xu, Gaoyuan Du, Xiangting Jia, Dongmin Zhao, Chunyan Zhou, Kexin Xiao, Jia Gu, Junnan Zhu, Xuequn Shang   +9 more
wiley   +1 more source