What question did this study set out to answer?

The aim is to quantify how molecular features influence cell phenotype using a new computational method.

April 25, 2026Open Access

Geometry-aware graph attention networks to explain single-cell chromatin states and gene expression with SEAGALL

Key Points

The aim is to quantify how molecular features influence cell phenotype using a new computational method.
Developed SEAGALL, a pipeline integrating geometry-regularised autoencoders with explainable graph attention networks.
Utilized a deep learning framework to embed single-cell data into a latent space for accurate cell-cell graph construction.
Applied explainable artificial intelligence techniques to interpret the predictions and highlight key features affecting cell identity.
SEAGALL identified stable signatures across diverse omics experiments, enhancing understanding of cell identity.
The method provides clear explanations of the features influencing chromatin states and gene expression.

Abstract

Abstract High-throughput single-cell sequencing is widely used to study cell identity. We present SEAGALL (Single-cell Explainable Geometry-Aware Graph Attention Learning pipeLine), a deep learning method to quantify the impact of molecular features on cellular phenotype, based on geometry-regularised autoencoders (GRAE) and explainable graph attention networks (X-GAT). The GRAE embeds the data into a latent space to build a reliable cell-cell graph. The GAT is trained to learn the annotations and XAI is used to explain the predictions, unravelling the features driving cell identity. SEAGALL extracts specific and stable signatures from multiple omics experiments, going beyond differential marker genes.

Bookmark

View Full Paper

Cite This Study

Malagoli et al. (Thu,) studied this question.

synapsesocial.com/papers/69ec5a2588ba6daa22dabbd5 https://doi.org/https://doi.org/10.1186/s13059-026-04066-2

Bookmark

View Full Paper