Unlocking Tissue Mechanics with TensionMap: Elevating Spatial Analysis

In this post, we’re thrilled to introduce TensionMap, a cutting-edge computational tool that takes your spatial transcriptomics data to the next level by revealing the mechanical forces shaping tissues.

What is TensionMap?

Published in Nature Methods (March 2025), TensionMap is a computational framework designed to infer mechanical forces acting on individual cells within tissues. Developed to work hand-in-hand with spatial transcriptomics data, TensionMap bridges the gap between molecular profiling and the physical environment of cells. While traditional spatial transcriptomics excels at mapping gene expression, it often overlooks the mechanical cues—like tension and pressure—that influence cell behavior and tissue patterning. TensionMap fills this gap by leveraging cell segmentation results to estimate forces and provide a richer understanding of tissue architecture.

Think of TensionMap as a translator: it takes the spatial coordinates and morphological data from segmented tissue images and converts them into a "force map" that reveals how cells push, pull, and interact mechanically. This is especially valuable in dynamic processes like embryonic development, wound healing, or cancer progression, where mechanical feedback plays a pivotal role alongside biochemical signals.

How TensionMap Works with Cell Segmentation

TensionMap builds on the foundation laid by cell segmentation methods, which we discussed in our last post. Whether you’re using deep learning-based tools like Cellpose or segmentation-free approaches like SSAM, TensionMap can take those outputs—cell boundaries, shapes, and positions—and analyze them to infer mechanical interactions. Here’s a simplified breakdown of how it works:

1. Input Data: TensionMap starts with spatially resolved transcriptomics data paired with segmented tissue images. The segmentation provides critical information about cell size, shape, and spatial arrangement.

2. Force Inference: Using a sophisticated computational model, TensionMap calculates the forces acting on each cell based on its morphology and its neighbors’ configurations.

3. Integration with Transcriptomics: By combining these mechanical insights with gene expression profiles, TensionMap identifies signatures—such as mechanical stress patterns or gene modules—that correlate with tissue boundaries or cellular states.

4. Output: The result is a detailed map of mechanical forces across the tissue, offering clues about how physical dynamics influence biological processes.

For example, in a developing mouse embryo (a case study from the Nature Methods paper), TensionMap used spatial transcriptomics data to pinpoint forces driving tissue compartment boundaries—insights that purely molecular data couldn’t fully explain. This dual lens of mechanics and transcriptomics opens up new avenues for discovery.

Why TensionMap Matters

We believe that the future of bioinformatics lies in integrating diverse data types to paint a complete picture of biological systems. TensionMap exemplifies this approach by adding a mechanical dimension to spatial transcriptomics. Here’s why it’s a game-changer:

• Enhanced Tissue Understanding: Mechanical forces shape everything from cell differentiation to tissue morphogenesis. TensionMap reveals these hidden drivers, complementing gene expression data.

• Applications Across Fields: Whether you’re studying developmental biology, cancer microenvironments, or regenerative medicine, TensionMap provides actionable insights into how tissues function and adapt.

• Seamless Integration: It works with existing spatial platforms (like Visium or MERFISH) and segmentation pipelines, making it a versatile addition to your analysis toolkit.