Inferring super-resolution tissue architecture by integrating spatial transcriptomics with histology.
Daiwei ZhangAmelia SchroederHanying YanHaochen YangJian HuMichelle Y Y LeeKyung S ChoKatalin SusztákGeorge X XuMichael D FeldmanEdward B LeeEmma E FurthLinghua WangMingyao LiPublished in: Nature biotechnology (2024)
Spatial transcriptomics (ST) has demonstrated enormous potential for generating intricate molecular maps of cells within tissues. Here we present iStar, a method based on hierarchical image feature extraction that integrates ST data and high-resolution histology images to predict spatial gene expression with super-resolution. Our method enhances gene expression resolution to near-single-cell levels in ST and enables gene expression prediction in tissue sections where only histology images are available.
Keyphrases
- gene expression
- single cell
- deep learning
- dna methylation
- high resolution
- rna seq
- convolutional neural network
- induced apoptosis
- high throughput
- machine learning
- cell cycle arrest
- mass spectrometry
- single molecule
- artificial intelligence
- big data
- electronic health record
- signaling pathway
- climate change
- risk assessment
- human health