SpaGene: A Deep Adversarial Framework for Spatial Gene Imputation.

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) offers high-resolution gene expression but lacks spatial context.
• Spatial transcriptomics provides spatial resolution but has limited transcriptomic coverage.
• Integrating both data types is crucial for comprehensive tissue analysis.

Purpose of the Study:

• To introduce SpaGene, a novel deep learning framework for integrating scRNA-seq and spatial transcriptomics data.
• To impute missing gene expressions in spatial transcriptomics datasets using scRNA-seq data.
• To enhance the biological insights derived from spatial transcriptomics.

Main Methods:

• SpaGene utilizes a deep learning architecture with encoder-decoder pairs, translators, and discriminators.
• The framework integrates transcriptome-wide single-cell gene expression data with spatial context.
• Performance was benchmarked against state-of-the-art methods across diverse datasets.

Main Results:

• SpaGene achieved superior performance, with an average 33% higher Pearson correlation coefficient (PCC), 21% higher Structural Similarity Index (SSIM), and 6.6% lower Root Mean Squared Error (RMSE).
• The model reliably imputes missing genes, providing comprehensive transcriptomic profiles.
• Application to lung tumor tissue revealed immune cell enrichment at tumor boundaries and restricted myeloid cell trafficking.

Conclusions:

• SpaGene effectively integrates scRNA-seq and spatial transcriptomics data, enhancing spatial transcriptomics capabilities.
• The framework provides spatially resolved, enhanced transcriptome data for deeper biological understanding.
• Findings offer novel insights into tumor-immune interactions and potential therapeutic development avenues.