graph neural network

Reference-free variant calling with local graph construction with ska lo (SKA)

This paper introduces ska lo, a reference-free variant calling tool that uses a coloured De Bruijn graph to detect SNPs, insertions, deletions, and recombination events in pathogen genomes.

HRCHY-CytoCommunity identifies hierarchical tissue organization in cell-type spatial maps

HRCHY-CytoCommunity identifies hierarchical multicellular structures in spatial cell-type maps and improves patient stratification in breast cancer.

MetDeeCINE: Deciphering Metabolic Regulation through Deep Learning and Multi-Omics

MetDeeCINE predicts enzyme-metabolite relationships using multi-omics data and outperforms existing models in predicting concentration control coefficients in E. coli.

SpaBatch: Batch Alignment of Spatial Transcriptomics Data using Graph Deep Learning

SpaBatch aligns spatial transcriptomics data across batches using graph deep learning to reduce technical variability and enable multi-slice integration.

Gauge equivariant convolutional neural networks for diffusion mri

This paper introduces gauge equivariant convolutional neural networks for diffusion MRI to improve angular resolution with fewer gradient directions.

Downregulation of Keratin Proteins in the Neuroendocrine System of Amyotrophic Lateral Sclerosis with PT150 as a Dual Neuroprotectant

The paper uses a Graph Neural Network to diagnose ALS with high accuracy and identifies keratin proteins and PT150 as potential biomarkers and therapeutics.

Contextualizing gene expression with feature rich graph neural networks

This paper introduces OGL, a deep learning method that integrates multi-omics data to predict gene expression and uncover regulatory interactions across cell lines and tissues.

Combining transformer and 3DCNN models to achieve co-design of structures and sequences of antibodies in a diffusional manner

The paper introduces AlphaPanda, an algorithm combining transformer, 3DCNN, and diffusion models for antibody sequence and structure design, achieving high-quality results with reduced training data.

Segger: Fast and accurate cell segmentation of imaging-based spatial transcriptomics data

Segger improves cell segmentation in spatial transcriptomics by using a graph neural network, achieving high sensitivity and speed with minimal manual intervention.

Frequency-Specific Resting-State MEG Network Characteristics of Tinnitus Patients Revealed by Graph Learning

This study uses graph learning to analyze resting-state MEG data and identifies frequency-specific functional connectome differences in tinnitus patients compared to controls.

Fiora: Local neighborhood-based prediction of compound mass spectra from single fragmentation events

Fiora predicts compound mass spectra from single fragmentation events using graph neural networks, outperforming existing methods and enabling faster spectral library expansion.

Exploring Comorbidity Networks in Mild Traumatic Brain Injury Subjects through Graph Theory: A Traumatic Brain Injury Model Systems Study

This study uses graph theory to analyze comorbidity networks in mild TBI patients, revealing age-dependent associations between psychiatric and non-psychiatric conditions.

Encoding Drug-Target-Pathway-Disease profiles for Drug-Cancer Association Prediction using Graph Transformer-Convolution Networks

The paper proposes a Graph Transformer-Convolution Network (GTCN) to predict drug-cancer associations by integrating heterogeneous biological data and enhancing interpretability.

Mapping the landscape of social behavior

This paper introduces a 3D tracking technique to analyze social behavior in rodents, revealing new insights into autism and amphetamine effects through high-resolution postural data.

A major disease-related point mutation in spastin alters dramatically the dynamics and allostery of the motor

This study uses molecular dynamics and machine learning to analyze how the R591S mutation in spastin affects its dynamics and allostery, leading to impaired microtubule severing and HSP-related dysfunction.

SCOIGET: Predicting Spatial Tumor Evolution Pattern by Inferring Spatial Copy Number Variation Distributions

SCOIGET infers spatial copy number variation distributions using graph neural networks, improving accuracy and efficiency in tumor evolution analysis.

AI-Driven Breeding Enhances Stress Tolerance in High-Elevation Extremophytes: A Proof-of-Concept Study with Cross-Component Validation

This study demonstrates an AI platform that predicts and enhances stress tolerance in high-elevation extremophytes using multi-omics data and machine learning techniques.

Local graph-motif features improve gene interaction network prediction

This paper introduces LAMP features to enhance VGAEs for gene interaction network prediction, showing improved performance on missing-link prediction tasks.

Evolutionary rewiring of the dynamic network underpinning allosteric epistasis in NS1 of influenza A virus

This study analyzes how mutations in the NS1 protein of influenza A virus rewired allosteric and epistatic networks, enhancing its binding affinity to p85β.

stDyer enables spatial domain clustering with dynamic graph embedding

stDyer improves spatial domain clustering in SRT data by combining GMVAE and GATs with dynamic graphs and mini-batch sampling, outperforming existing methods.

Enhanced identification of small molecules binding to hnRNPA1 via cryptic pockets mapping coupled with X-Ray fragment screening

This study identifies small molecule binders to hnRNPA1 using MD simulations and fragment screening, revealing cryptic pockets and conformational changes.

Understanding the link between functional profiles and intelligence through dimensionality reduction and graph analysis

This study links functional connectivity profiles to fluid intelligence using dimensionality reduction and graph analysis, showing that functional distance from sensory and high-cognition systems correlates with intelligence.

Sidechain conditioning and modeling for full-atom protein sequence design with FAMPNN

This paper introduces FAMPNN, a method for protein sequence design that explicitly models both amino acid identity and sidechain conformation, achieving state-of-the-art results in sidechain packing and practical design applications.

How to deal with internal fragment ions?

The paper introduces a graph-based method to annotate internal fragment ions in tandem mass spectra, improving proteoform sequencing and modification localization.

Predicting mutation-disease associations through protein interactions via deep learning

This paper constructs a mutation-disease dataset and proposes CGM to predict mutation-disease associations, identifying 22 unknown protein interaction pairs.

GraphMatch: Knowledge Graphs for Allogeneic Stem Cell Matching

GraphMatch is a scalable graph database for allogeneic stem cell matching, demonstrating efficient search performance with large datasets.

M3NetFlow: A novel multi-scale multi-hop graph AI model for integrative multi-omic data analysis

M3NetFlow is a multi-scale multi-hop graph model for multi-omic data analysis that ranks targets and infers signaling pathways, achieving high accuracy in drug response and Alzheimer's disease studies.

An intricate balancing act: Upstream and downstream frameshift co-regulatory elements

The paper investigates the RNA structures and interactions of the 5' and 3' regions of the SARS-CoV-2 frameshifting element, revealing their co-regulatory roles and implications for therapeutic interventions.

Boosting protein graph representations through static-dynamic fusion

The paper introduces a graph representation that combines static and dynamic protein data, improving performance in three tasks related to protein-ligand interactions.

Image-based spatial transcriptomics identifies molecular niche dysregulation associated with distal lung remodeling in pulmonary fibrosis

Xenium profiling of 343 genes in 1,126,820 nuclei from 19 human lungs maps fibrosis niches and orders alveolar remodeling from alveolar-capillary disruption to SPP1+ macrophage accumulation.