graph neural network

Representational learning of brain responses in executive function and higher-order cognition using deep graph convolutions

This paper uses deep graph convolutions to decode brain activity across cognitive domains, finding that integration boosts high-order cognition decoding while segregation suffices for motor tasks.

To knot or not to knot: Multiple conformations of the SARS-CoV-2 frameshifting RNA element

The paper identifies multiple structural conformations of the SARS-CoV-2 frameshifting RNA element and their potential roles in viral replication and therapeutic targeting.

Fast and effective protein model refinement by deep graph neural networks

This paper introduces a fast protein model refinement method using graph neural networks, achieving high quality results with limited conformation sampling.

Protein Interaction Network-based Deep Learning Framework for Identifying Disease-Associated Human Proteins

The paper introduces PINDeL, a GCN-based model for identifying disease-associated human proteins, achieving high accuracy and outperforming existing methods with experimentally verified results.

Learning cell communication from spatial graphs of cells

The paper introduces NCEM, a graph neural network method for modeling cell communication in spatial molecular profiling data, identifying niche-dependent gene expression and cell state variation across multiple tissues.

Circuit motifs and graph properties of connectome development in C. elegans

The paper analyzes C. elegans connectomes across developmental stages using graph and motif analysis to identify evolving network properties and stable patterns.

scSGL: Signed Graph Learning for Single-Cell Gene Regulatory Network Inference

scSGL infers gene regulatory networks from single-cell RNA sequencing data by learning signed graphs that capture both activating and inhibitory gene relationships, while handling cell-cycle heterogeneity and dropouts.

Single-Cell Classification Using Graph Convolutional Networks

This study introduces a deep learning model called sigGCN that combines graph convolutional networks with gene expression data to improve cell type classification accuracy.

APRILE: Exploring the Molecular Mechanisms of Drug Side Effects with Explainable Graph Neural Networks

APRILE uses explainable graph neural networks to identify molecular mechanisms of polypharmacy side effects, generating explanations for over 900,000 side effect-drug pairs.

Mako: a graph-based pattern growth approach to detect complex structural variants

Mako detects complex structural variants using a graph-based pattern growth approach, achieving 70% validation rate on real data and identifying 15 CSV types including two novel ones.

Discovering cancer driver genes and pathways using stochastic block model graph neural networks

This paper introduces SBM-GNN, a graph neural network model that identifies cancer driver genes and pathways from omic data, outperforming existing methods and revealing key cancer-related processes.

Connectome-based individual prediction of cognitive behaviors via the graph propagation network reveals directed brain network topology

This study introduces a graph propagation network to predict cognitive behaviors from brain connectomes, finding it outperforms traditional methods by leveraging directed network topology.

Probing the rules of cell coordination in live tissues by interpretable machine learning based on graph neural networks

The paper uses graph neural networks to infer cell coordination rules from live tissue data, revealing region-specific coordination patterns in mammalian epidermis.

scGCN: a Graph Convolutional Networks Algorithm for Knowledge Transfer in Single Cell Omics

scGCN transfers labels across single-cell omics datasets with superior accuracy by using a graph-based AI model.

DotMotif: An open-source tool for connectome subgraph isomorphism search and graph queries

DotMotif is an open-source tool that enables efficient subgraph isomorphism search and graph queries in connectomics data, demonstrated on simulated and real datasets.

X-DPI: A structure-aware multi-modal deep learning model for drug-protein interactions prediction

X-DPI is a structure-aware multi-modal deep learning model for drug-protein interactions prediction that outperforms existing methods on benchmark datasets.

Bridging brain and cognition: A multilayer network analysis of brain structural covariance and general intelligence in a developmental sample of struggling learners

The paper uses network analysis to model brain structural covariance and cognitive abilities in struggling learners, finding positive correlations and intermediary roles of specific nodes.

Brain network dynamics codify heterogeneity in seizure propagation

This study uses graph-theoretical analysis to identify network dynamics predicting seizure propagation patterns and post-onset dynamics in epilepsy.

Exploiting node metadata to predict interactions in large networks using graph embedding and neural networks

The paper presents a method combining graph embedding and machine learning to predict network interactions using node metadata, showing it works well on a tourist visit dataset.

MORONET: Multi-omics Integration via Graph Convolutional Networks for Biomedical Data Classification

MORONET is a multi-omics integration method using graph convolutional networks for biomedical classification, outperforming existing approaches and identifying disease-related biomarkers.

SE(3)-Equivariant Energy-based Models for End-to-End Protein Folding

This paper introduces SE(3)-Fold, an end-to-end protein folding method using SE(3)-equivariant energy-based models, achieving comparable accuracy to state-of-the-art methods with significant speed improvements.

NeTOIF: A Network-based Approach for Time-Series Omics Data Imputation and Forecasting

NeTOIF improves time-series omics data imputation and forecasting by integrating network-based methods and LSTM networks, outperforming existing techniques in handling missing values across multiple time points.

HGC: fast hierarchical clustering for large-scale single-cell data

HGC enables efficient multiresolution clustering of large-scale single-cell data by combining graph-based and hierarchical clustering methods, achieving state-of-the-art accuracy.

PIGNON: A protein-protein interaction-guided functional enrichment analysis for quantitative proteomics

PIGNON detects differentially expressed functional annotations by mapping protein expression onto a protein-protein interaction network, identifying 168, 203, and 19 breast cancer pathways between and within breast cancer subtypes.

Deciphering biological evolution exploiting the topology of Protein Locality Graph

This paper introduces a method using Protein Locality Graph topology to infer species trees by analyzing inter-module interactions in protein networks.

Distillation of MSA Embeddings to Folded Protein Structures with Graph Transformers

The paper introduces a graph transformer architecture that uses MSA embeddings to predict protein structures directly from sequences.

Using Graph Convolutional Neural Networks to Learn a Representation for Glycans

This paper introduces SweetNet, a graph convolutional neural network for glycan sequence representation, showing it outperforms existing methods and predicts organismal traits and viral glycan-binding.

A Biologically Interpretable Graph Convolutional Network to Link Genetic Risk Pathways and Neuroimaging Phenotypes of Disease

This paper introduces a graph convolutional network that links genetic risk pathways with neuroimaging phenotypes in schizophrenia, achieving better classification performance than existing methods.

Structure-Based Function Prediction using Graph Convolutional Networks

This paper introduces a Graph Convolutional Network for predicting protein function and identifying important residues, showing improved accuracy over sequence-based methods.

Leveraging Sequential and Spatial Neighbors Information by Using CNNs Linked With GCNs for Paratope Prediction

The paper introduces a method combining CNNs and GCNs to predict paratope residues in antibodies by integrating sequence and spatial information.