MAD: Microenvironment-Aware Distillation -- A Pretraining Strategy for Virtual Spatial Omics from Microscopy

2026-03-11

AI Summary

Current omics technologies struggle to simultaneously achieve single-cell resolution and tissue-scale molecular profiling due to limitations in cost and throughput. To address this challenge, this work proposes a microenvironment-aware, dual-view self-supervised pretraining method that jointly leverages self-distillation between the morphological views of individual cells and their local microenvironments to construct a cell-centric unified embedding representation. The approach requires no large-scale annotated data and is applicable across diverse tissue types and imaging modalities. In downstream tasks—including cell subtype classification, transcriptome prediction, and biological inference—the model outperforms foundation models with comparable parameter counts that rely on substantially more training data, significantly enhancing the efficiency and generalization capability of in situ single-cell analysis in tissues.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Computer Vision: Medical and Biological Imaging
• Search and Optimization: Learning to Search

Application Category

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

Abstract

Bridging microscopy and omics would allow us to read molecular states from images-at single-cell resolution and tissue scale-without the cost and throughput limits of omics technologies. Self-supervised pretraining offers a scalable approach with minimal labels, yet how to encode single-cell identity within tissue environments-and the extent of biological information such models can capture-remains an open question. Here, we introduce MAD (microenvironment-aware distillation), a pretraining strategy that learns cell-centric embeddings by jointly self-distilling the morphology view and the microenvironment view of the same indexed cell into a unified embedding space. Across diverse tissues and imaging modalities, MAD achieves state-of-the-art prediction performance on downstream tasks including cell subtyping, transcriptomic prediction, and bioinformatic inference. MAD even outperforms foundation models with a similar number of model parameters that have been trained on substantially larger datasets. These results demonstrate that MAD's dual-view joint self-distillation effectively captures the complexity and diversity of cells within tissues. Together, this establishes MAD as a general tool for representation learning in microscopy, enabling virtual spatial omics and biological insights from vast microscopy datasets.

Problem

• microenvironment
• single-cell identity
• virtual spatial omics
• self-supervised pretraining
• morphology

Innovation

• microenvironment-aware distillation
• self-supervised pretraining
• joint self-distillation
• virtual spatial omics
• cell-centric embedding