GC-MoE: Genomics-Guided Cell-Type-Specific Mixture of Experts for Histology-Based Single-Cell Spatial Transcriptomics
2026-06-01 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionArtificial IntelligenceMachine Learning
AI summaryⓘ
The authors developed a new method called GC-MoE to predict gene activity in individual cells using microscope images and cell locations. Unlike previous methods that look at groups of cells, their approach focuses on differences between single cells based on their type. They use special components that learn patterns specific to each cell type and consider signals from neighboring cells to improve accuracy. Testing their method on public data showed better results than existing approaches.
single-cell spatial transcriptomicshistopathological imagesgene expression predictioncell typemixture-of-expertscell-to-cell interactionco-expressionrouting networkattention mechanism
Authors
Kaito Shiku, Ahtisham Fazeel Abbasi, Ryoma Bise, Yuichiro Iwashita, Kazuya Nishimura, Andreas Dengel, Muhammad Nabeel Asim
Abstract
Histology-based single-cell spatial transcriptomics (ST) estimation aims to predict gene expression for individual cells from histopathological images and cell locations, reducing the need for costly single-cell ST measurements. Unlike existing histology-to-ST methods that mainly predict spot-level profiles for local regions containing multiple cells, this task requires modeling cell-to-cell expression variability, which is strongly structured by cell type. We propose Genomics-Guided Cell-Type-Specific Mixture-of-Experts (GC-MoE), which estimates cell-type probabilities with a routing network and softly combines cell-type-specific experts for gene expression prediction. To further encode cell-type-dependent gene programs, we introduce the Cell-Type-Specific Co-Expression-Aware Predictor (CAP), together with a lightweight Cell-to-Cell Interaction Attention (C2CA) module for neighboring-cell context. Experiments and ablations on public single-cell ST datasets show consistent improvements over existing single-cell and adapted spot-level baselines.