graph neural network

Physics-Diffusion-Driven Multiscale Aggregation for Drug-Target Interaction Prediction

PDDMA-DTI improves drug-target interaction prediction by combining physics-inspired diffusion with multiscale aggregation and interaction enhancement, achieving better accuracy and robustness than existing methods.

Atlas-independent brain connectome analysis at voxel-level granularity: graph convolutional networks for etiology classification in newborns

This paper uses graph convolutional networks to classify neonatal brain connectomes by etiology at voxel-level granularity, achieving higher classification performance than atlas-based methods and a multilayer perceptron baseline.

SynOmics: Integrating Multi-omics Data Through Feature Interaction Networks

SynOmics improves multi-omics integration by modeling feature interactions through a graph convolutional network, outperforming existing methods in biomedical classification tasks.

Improved Graph-based Antibody-aware Epitope Prediction with Protein Language Model-based Embeddings

This paper improves epitope prediction using graph convolutional networks and protein language model embeddings, achieving an AUCROC of 0.65 on a benchmark dataset.

ProAffinity++

ProAffinity++ predicts protein-protein binding affinities using sequence and structure representations, outperforming existing methods in accuracy and handling complex cases like antigen-antibody interactions.

Revisiting graph-based approaches for small protein analysis: Insights from anti-CRISPR protein networks

The paper evaluates graph-based methods for predicting functional residues in small anti-CRISPR proteins, finding that existing techniques fail due to structural biases in network representations.

Tile-X: A vertex reordering approach for scalable long read assembly

Tile-X improves long read assembly by precomputing read ordering, enhancing scalability and efficiency without sacrificing quality.

Decision Voting Based Multiscale Convolutional Learning of Brain Networks With Explainability

This paper introduces a multiscale GCN framework for brain network analysis that integrates structural and functional connectivity data, achieving 71.59% accuracy in schizophrenia classification with explainable insights.

Informational and methodological differences in regional structure-function coupling in modeling approaches

The paper compares four modeling approaches for calculating regional structure-function coupling in the brain, revealing how indirect connections affect their estimations.

NetFlow: A Framework to Explore Topological Representations of High-Dimensional Biomedical Data

NetFlow constructs graph representations of biomedical data to reveal structured patterns and refine clinical subtypes in cancer datasets.

Deep learning for mass extinction detection on fossilized phylogenies: power, limitations, and lessons for simulation-based phylodynamic inference

This study evaluates a deep learning model for detecting mass extinctions in fossilized phylogenies, finding it slightly better than Bayesian methods but limited by simulation stochasticity and definition specificity.

SLOGEN: A Structure-based Lead Optimization Model Unifying Fragment Generation and Screening

Slogen unifies fragment generation and screening for lead optimization, improving chemical diversity and synthetic tractability while matching state-of-the-art performance.

Hypercholesterolemia Risk Prediction from Serum Metabolomics Using a Metabolic Pathway-Integrated Graph Neural Network

This study introduces MP-GNN, a graph neural network integrating metabolic pathways to predict hypercholesterolemia risk from serum metabolomics data, achieving high accuracy and biological interpretability.

Structured Chemical Reaction Modeling with Multitask Graph Neural Networks

This paper introduces a multitask graph neural network for chemical reaction modeling that jointly predicts atom mappings, reaction centers, and classifications, achieving state-of-the-art results on a benchmark.

Improved Spatial Transcriptomics Clustering with Nested Graph Neural Networks

STING improves spatial transcriptomics clustering by integrating gene-gene and spatial relations through nested graph neural networks, achieving better clustering performance across multiple datasets.

High-precision cell-type mapping and annotation of single-cell spatial transcriptomics with STAMapper

STAMapper maps and annotates cell types in single-cell spatial transcriptomics data using a graph neural network, achieving high accuracy and enabling unknown cell-type detection.

Temporal Expression Prediction by Integrating Genome Dynamics via Spatio-temporal GNNs

The paper introduces spatio-temporal GNNs to predict mRNA and miRNA expression over time using Hi-C data, achieving high correlation and variance explanation.

Learning universal knowledge graph embedding for predicting biomedical pairwise interactions

This paper introduces LukePi, a self-supervised pre-training framework for GNNs on biomedical knowledge graphs, which improves link prediction in low-data and distribution-shift scenarios.

Identifying tissue states by spatial protein patterns related to chemotherapy response in triple-negative breast cancer

This study develops a computational pipeline for analyzing spatial proteomics data to identify tissue states associated with chemotherapy response in TNBC, revealing distinct spatial patterns and predictive markers.

dFCExpert: Learning Dynamic Functional Connectivity Patterns with Modularity and State Experts

dFCExpert learns dynamic functional connectivity patterns from fMRI data using modularity and state experts, showing superior performance on large-scale datasets.

RNA sequence design and protein-DNA specificity prediction with NA-MPNN

This paper introduces NA-MPNN, a unified deep learning model for nucleic acid inverse folding, which outperforms existing methods in RNA sequence design and protein-DNA specificity prediction.

Unlocking data in Klebsiella lysogens to predict capsular type-specificity of phage depolymerases

This paper models phage depolymerase capsular tropism using graph neural networks and sequence clustering, showing predictive ability and transferability to lytic phages.

Peptide Design through Binding Interface Mimicry

PepMimic, an all-atom latent diffusion model guided by a contrastive interface encoder, designs peptides mimicking known binders; 8% of tested peptides across five targets reached 10^-8M KD.

Integrative Spatial Modelling of Cellular Plasticity using Graph Neural Networks and Geostatistics

The paper introduces a framework combining graph neural networks and geostatistics to predict and explain cell state variation in spatial transcriptomics data, showing mesenchymal states are stabilized by genetic and environmental factors while hybrid states remain plastic.

Exploring Augmentation-Driven Invariances for Graph Self-supervised Learning in Spatial Omics

This study evaluates how different graph augmentations affect embedding quality and downstream performance in spatial omics, showing that well-designed augmentations improve results while poor ones can harm performance.

HiC-SCA: A Spectral Clustering Method for Reliable A/B Compartment Assignment From Hi-C Data

HiC-SCA improves A/B compartment assignment from Hi-C data through spectral clustering with noise filtering and resolution selection, showing better consistency across datasets.

Multimodal Cross-Attentive Graph-Based Framework for Predicting In Vivo Endocrine Disruptors

The paper presents a framework that predicts endocrine disruptors using a multimodal cross-attentive graph model, achieving high accuracy and providing mechanistic explanations.

Anticipating invasion of parasite with graph theory: Dryocosmus kuriphilus, a threat for Castanea sativa

The paper models the spread of the chestnut gall wasp using graph theory to assess invasion risk based on chestnut tree distribution.

MORGaN: self-supervised multi-relational graph learning for drug target discovery

MORGaN is a self-supervised framework for drug target discovery that outperforms existing models by leveraging multi-omic, multi-relation gene networks and provides interpretable results.

Combinatorial prediction of therapeutic perturbations using causally-inspired neural networks

PDGrapher predicts combinatorial therapeutic targets to reverse disease phenotypes using a causally inspired GNN, outperforming existing methods in cell line and genetic perturbation datasets.