AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

SAMGTD: Spatial-Aware Masked Graph Transformer-Diffusion Model for Enhanced Cell Type Deconvolution in Spatial Transcriptomics

Shilin Zhang, Suixue Wang, Qingchen Zhang, Xiulong Liu

AAAI 2026 · Graph learning › Graph self-supervised learning · approach: Brain and biological graph learning · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add spatial diffusion module for denoising and imputation (ablation)SAMGTD · Human lymph node (B GC prePB)AUC: 0.9116 -> 0.9595Improved
add spatial diffusion module (ablation)SAMGTD · Human lymph node (B GC DZ)AUC: 0.9294 -> 0.9761Improved
add masked graph transformer module (ablation)SAMGTD · Human lymph node (B GC DZ)AUC: 0.9121 -> 0.9761Improved
add masked graph transformer module (ablation)SAMGTD · Human lymph node (B Cycling)AUC: 0.9269 -> 0.9662Improved
add VAE denoising for scRNA-seq (ablation)SAMGTD · Human lymph node (B GC prePB)AUC: 0.9462 -> 0.9595Improved
add spatial-aware contrastive learning (ablation)SAMGTD · Human lymph node (B Cycling)AUC: 0.9657 -> 0.9662No clear effect
add spatial-aware contrastive learning (ablation)SAMGTD · Human lymph node (B GC DZ)AUC: 0.9701 -> 0.9761Improved
use SAMGTD instead of cell2location and GraphSTcell type deconvolution · Human lymph node (4 germinal-center cell types)AUCImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy