Spatial Transcriptomics Analysis of Spatially Dense Gene Expression Prediction

2025-03-03

AI Summary

Current spatial transcriptomics methods map image-based spots to single gene expression vectors, resulting in loss of spatial resolution—due to cellular heterogeneity within spots—and fixed-scale analysis. This work proposes PixNet, the first end-to-end deep learning model that generates dense, pixel-level gene expression maps, breaking the conventional “spot-to-vector” paradigm to enable continuous, high-resolution expression prediction. PixNet jointly leverages image semantic segmentation and spatial-expression disentangled modeling to support region-wise expression inference for spots of arbitrary size and across multiple scales—from subcellular to tissue-level. Evaluated on three benchmark datasets, PixNet significantly outperforms state-of-the-art methods, achieving consistent improvements in cross-scale prediction accuracy.

Technology Category

• Computer Vision: Representation Learning for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Natural Language Processing: Generation

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Spatial transcriptomics (ST) measures gene expression at fine-grained spatial resolution, offering insights into tissue molecular landscapes. Previous methods for spatial gene expression prediction usually crop spots of interest from pathology tissue slide images, and learn a model that maps each spot to a single gene expression profile. However, it fundamentally loses spatial resolution of gene expression: 1) each spot often contains multiple cells with distinct gene expression; 2) spots are cropped at fixed resolutions, limiting the ability to predict gene expression at varying spatial scales. To address these limitations, this paper presents PixNet, a dense prediction network capable of predicting spatially resolved gene expression across spots of varying sizes and scales directly from pathology images. Different from previous methods that map individual spots to gene expression values, we generate a dense continuous gene expression map from the pathology image, and aggregate values within spots of interest to predict the gene expression. Our PixNet outperforms state-of-the-art methods on 3 common ST datasets, while showing superior performance in predicting gene expression across multiple spatial scales. The source code will be publicly available.

Problem

• Predicts gene expression at varying spatial scales.
• Overcomes loss of spatial resolution in existing methods.
• Generates dense continuous gene expression maps from images.

Innovation

• PixNet predicts gene expression across varying spatial scales.
• Generates dense continuous gene expression maps from pathology images.
• Aggregates gene expression values within spots of interest.