graph neural network

ModelAngelo combines convolutional and graph neural networks with protein language model embeddings and hidden Markov model searches to automate atomic model building and identify unknown proteins in cryo-EM maps.

This paper introduces a graph diffusion autoregressive model to estimate dynamic functional connectivity from neural field potentials, validated on simulated data and macaque recordings.

MaskGraphene integrates multi-slice, multi-condition spatial transcriptomics data using a graph neural network to generate interpretable joint embeddings with preserved geometric information, improving downstream analyses.

The paper introduces NicheCompass, a graph deep learning method for characterizing cell niches in spatial omics data, showing improved performance in identifying and analyzing niches across multiple biological contexts.

This study uses 3D CNN models to classify mitosis stages in 3D time-lapse cell sequences, achieving higher accuracy than previous methods.

GNASTy improves pneumococcal serotyping by using a pangenome graph algorithm with Nanopore Adaptive Sampling, enhancing sensitivity for novel serotypes in complex samples.

EGGNet is a geometric deep learning framework for predicting molecule-protein interactions, achieving competitive performance on binding affinity and interface prediction tasks.

TopEC improves enzyme function classification using 3D graph neural networks and a localized 3D descriptor, achieving high accuracy in EC class prediction.

This paper introduces a graph convolutional network to predict disease-gene associations by leveraging external collaborative graphs and reducing noise through mutual information maximization.

CellDiffusion improves scRNA-seq data integration by accurately inferring cell-cell associations and correcting batch effects across diverse datasets.

The paper introduces MEGA, a tool to identify cancer-associated microbes using RNA-seq data from the ORIEN dataset, leveraging graph attention networks and microbial community information.

This paper introduces XP-GNN, a method for explaining Graph Neural Networks in biological knowledge graphs by leveraging community structures, showing improved relevance in drug discovery use cases.

DeepVelo estimates cell-specific transcriptional kinetics in scRNA-seq data using GCNs, enabling analysis of complex differentiation processes with time-varying rates.

This paper introduces a graph neural network model to predict Gibbs free energy for metabolic reactions and identifies thermodynamic driver reactions in human metabolic networks.

Cellograph uses graph neural networks to analyze multi-condition single-cell RNA-seq data, quantifying perturbation effects at single-cell resolution and outperforming existing methods.

scDirect identifies key transcription factors driving cell state transitions using single-cell multi-omics data, showing robust performance in benchmarking and real-world applications.

MarsGT identifies rare cell populations in multi-omics single-cell data, outperforming existing tools and revealing novel biological insights in human and mouse datasets.

The paper introduces SNOWFLAKE, a graph neural network framework that predicts disease status from multiplex imaging data in human B-cell follicles, showing improved performance over other methods.

This paper introduces GEFMAP, a geometric deep learning method to predict metabolic flux from scRNA-seq data, estimating reaction rates in metabolic networks.

The paper introduces DeepDynaTree, a phylogenetic-informed graph deep learning model to classify dynamic transmission clusters in infectious disease epidemics, achieving higher accuracy than existing methods.

The paper uses graph neural networks to predict functional connectivity from structural connectivity in human brain networks, finding strong group-common structure-function coupling with some individual-specific effects.

The paper uses multiplexed imaging and graph convolutional networks to analyze mTOR signaling heterogeneity in colorectal tumors and its correlation with patient prognosis.

This paper introduces DeepRank-GNN-esm, a graph neural network that uses ESM-2 embeddings to score protein-protein models, replacing computationally expensive PSSM features and improving performance on docking and crystal artifact detection tasks.

The paper develops a model combining tumor-specific methylation atlas and cfDNA features to improve tissue-of-origin detection in low-depth cfDNA samples.

GeNNius uses a graph neural network to predict drug-target interactions more efficiently and accurately than existing methods across multiple datasets.

GSDensity provides pathway-centric analysis of single-cell and spatial transcriptomics data without clustering, revealing novel cell-pathway associations and spatially relevant pathways in cancer and brain development.

This paper predicts brain age in mouse AD models using FAGNN, identifying key neural connections and showing APOE2's neuroprotective role against aging.

This paper introduces a graph clustering algorithm for detecting and genotyping structural variants from long-read sequencing data, achieving high accuracy in benchmark tests.

HGNNPIP improves PPI prediction by addressing data bias through a hybrid graph neural network framework, outperforming existing methods on benchmark datasets.

This paper introduces SPIN-CGNN, a protein design method using contact maps and graph neural networks, showing improved sequence recovery and structural accuracy compared to existing methods.