Characterization of the heterogeneity in SARS-CoV-2 fitness dynamics via graph representation learning. [PDF]
Wang Z +14 more
europepmc +1 more source
PREdicting LNP In Vivo Efficacy (PRELIVE) framework enables the prediction of lipid nanoparticle (LNPs) organ‐specific delivery through dual modeling approaches. Composition‐based models using formulation parameters and protein corona‐based models using biological fingerprints both achieve high predictive accuracy across multiple organs.
Belal I. Hanafy +3 more
wiley +1 more source
Discriminative graph regularized representation learning for recognition. [PDF]
Qi J, Xu R.
europepmc +1 more source
Permanent magnets derive their extraordinary strength from deep, universal electronic‐structure principles that control magnetization, anisotropy, and intrinsic performance. This work uncovers those governing rules, examines modern modeling and AI‐driven discovery methods, identifies critical bottlenecks, and reveals electronic fingerprints shared ...
Prashant Singh
wiley +1 more source
MCOAN: multimodal contrastive representation learning for cross-omics adaptive disease regulatory network prediction. [PDF]
Long J, Liu B, Li J, Zhao S.
europepmc +1 more source
Designing Asymmetric Memristive Behavior in Proton Mixed Conductors for Neuromorphic Applications
Protonic devices that couple ionic and electronic transport are demonstrated as bioinspired neuromorphic elements. The devices exhibit rubber‐like asymmetric memristive behavior with slow voltage‐driven conductance increase and rapid relaxation, enabling simplified read–write operation.
Nada H. A. Besisa +6 more
wiley +1 more source
AUTOENCODIX: a generalized and versatile framework to train and evaluate autoencoders for biological representation learning and beyond. [PDF]
Joas MJ +4 more
europepmc +1 more source
Conductance‐Dependent Photoresponse in a Dynamic SrTiO3 Memristor for Biorealistic Computing
A nanoscale SrTiO3 memristor is shown to exhibit dynamic synaptic behavior through the interaction of local electrical and global optical signals. Its photoresponse depends quantitatively on the conductance state, which evolves and decays over tunable timescales, enabling ultralow‐power, biorealistic learning mechanisms for advanced in‐memory and ...
Christoph Weilenmann +8 more
wiley +1 more source
SlideMamba: entropy-based adaptive fusion of GNN and Mamba for enhanced representation learning in digital pathology. [PDF]
Khan S +5 more
europepmc +1 more source
Trap‐Assisted Transport and Neuromorphic Plasticity in Lead‐Free 2D Perovskites PEA2SnI4
An artificial retina built from lead‐free layered perovskite (PEA)2SnI4 converts light input into a persistent photocurrent and sums successive flashes over time. Micro/nanocrystals integrated on electrodes act as synapse‐like pixels that perform temporal integration directly in hardware. This in‐sensor preprocessing merges detection and computation on
Ofelia Durante +17 more
wiley +1 more source

