Dataset Information

DGAT: A Dual-Graph Attention Network for Inferring Spatial Protein Landscapes from Transcriptomics

ABSTRACT: This dataset comprises spatial multi-omic profiling of malignant pleural mesothelioma (MPM) tissue, enabling simultaneous measurement of transcriptomic and protein expression at spatial resolution. To address the lack of spatially resolved proteomic measurements in standard spatial transcriptomics (ST) technologies, we generated this in-house MPM spatial transcriptomics and proteomics dataset and used it to train and evaluate DGAT (Dual-Graph Attention Network), a deep learning framework for imputing protein expression from ST data. DGAT integrates transcriptomic, proteomic, and spatial information using graph attention networks and jointly reconstructs mRNA and protein profiles through multi-task learning. This dataset serves both as a biological resource for understanding the tumor immune microenvironment in MPM and as a benchmark for spatial protein inference. DGAT’s predictions on this dataset enabled improved identification of immune phenotypes and tumor–stroma spatial organization, with implications for biomarker discovery and therapeutic targeting in mesothelioma. Spatial transcriptomics (ST) technologies provide genome-wide mRNA profiles in tissue context but lack direct protein-level measurements, which are critical for interpreting cellular function and microenvironmental organization. We present DGAT (Dual-Graph Attention Network), a deep learning framework that imputes spatial protein expression from transcriptomics-only ST data by learning RNA–protein relationships from spatial CITE-seq datasets. DGAT constructs heterogeneous graphs integrating transcriptomic, proteomic, and spatial information, encoded using graph attention networks. Task-specific decoders reconstruct mRNA and predict protein abundance from a shared latent representation.

ORGANISM(S): Homo sapiens

PROVIDER: GSE300851 | GEO | 2026/04/22

REPOSITORIES: GEO

ACCESS DATA

Dataset's files

Source:

			Action	DRS
		Other

Items per page:

1 - 1 of 1

Similar Datasets

Project description:Spatial transcriptomics is a rapidly advancing field, yet it lacks standardized metrics for evaluating imaging-based in situ hybridization technologies. In this study, we emphasize the importance of rigorous experimental design and standardized operating procedures (SOPs) in defining performance metrics across multiple modalities—from single-cell to spatial transcriptomics to spatial proteomics. Our Spatial Touchstone (ST) dataset, generated from six tissue types across three global sites (n=76 assays), provides a comprehensive evaluation of reproducibility, sensitivity, dynamic ranges, signal-to-noise ratio, false discovery rates, cell type annotation, and congruence with single-cell profiling. Using single-nucleus RNA sequencing (snRNA-seq) on the same formalin-fixed paraffin-embedded tissues, we created a reference for transcriptional profiles to assess cellular organization and biomarker localization. In the absence of an independent 'gold standard’ in the field, this study provides an objective assessment of technical performance through standardized protocols, Spatial Touchstone Protocols (STSOP). Our open-source software, SpatialQC, allows users to evaluate samples across all technical metrics and directly impute cell annotations from single-cell datasets. This study features the largest imaging-based spatial transcriptomics data repository available, with a total of 203 spatial profiles. It incorporates both public (n=127) and ST datasets (n=76) through the Spatial Touchstone Portal (STP), available at www.spatialtouchstone.org. This resource offers users easy access to analyze and compare their own data against an extensive collection of imaging-based datasets. We show that the ST datasets demonstrate high reproducibility (r=0.93 to 1.0) and set a new benchmark for the field. Finally, we establish metrics to evaluate and integrate imaging based multi omics data from single cell into spatial transcriptomics to spatial proteomics. The ST project provides a comprehensive, reproducible assessment across platforms and sites, establishing essential standards for imaging-based spatial multi omics (transcriptomics and proteomics) and paving the way for future research and technological advancements.

Dataset Information

DGAT: A Dual-Graph Attention Network for Inferring Spatial Protein Landscapes from Transcriptomics

Dataset's files

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets