Hryhorii Chereda

This study benchmarks graph neural networks for cancer subtyping using methylation and gene expression data, finding that ChebNet performs best and simple readout layers improve performance.

The paper uses graph convolutional neural networks with layer-wise relevance propagation to analyze gene expression data and identify genes and pathways associated with BRCAness, outperforming traditional methods.

The paper evaluates the stability of feature selection using GCNN+LRP and GCNN+SHAP for gene expression data, comparing them to classical ML methods.

Ensemble-GNN enables federated, ensemble-based graph neural networks for disease module discovery using patient-specific PPI networks and omics data.

The paper adapts graph convolutional neural networks and graph layer-wise relevance propagation to the Spektral library for analyzing gene expression data from colorectal cancer patients, identifying relevant genes and subnetworks.

The paper introduces R-based tools for reproducible integration, exploration, and visualization of biological networks, with available code and documentation.

The paper introduces GLRP, a method to explain Graph-CNN decisions in breast cancer metastasis prediction, identifying patient-specific molecular subnetworks that align with clinical knowledge.

The paper develops word2vec-based embeddings from PubMed abstracts to capture biological relationships and improves metastatic event prediction using graph convolutional networks.