VIMA finds tissue patterns associated with disease—not just different cell counts
A new method learns recurring patterns in tissue patches, then tests whether their abundance differs between groups of samples.
Two tissues can contain similar mixtures of cells arranged differently.
VIMA, published in Nature Methods on October 6, addresses that spatial information: neural networks learn representations of small tissue patches; statistical tests ask which recurring patterns are associated with the condition under study. [1]
The authors applied it to existing datasets in rheumatoid arthritis, ulcerative colitis and dementia. The result is a map of associations worth investigating, rather than a diagnosis or an explanation of what caused the disease. Choosing a different patch size can change which patterns become visible. [1]
For a research group with spatial data, the useful next step is a comparison on its own samples. The public repository provides installation instructions and worked examples, including a read-only demonstration. Its interactive Colab example requires the paid GPU tier; Euangelion has inspected the documentation, not run the analysis. [2]
Sources & details 2 sources
Discuss in your voice app
Take the reporting, visual explanations and original sources into a conversation. Copy the brief, paste it into your preferred assistant, then start its voice mode.
Paste the brief into your voice app to discuss this reporting.Revised 9 October 2026. Copy edited 9 October to remove commentary about graphic design; reported claims, qualifications and source dates preserved.
Reporting sources
- Nature Methods · VIMA, October 6
Original open-access paper; methods, donor/sample counts, task-specific comparisons and limitations checked after collection.
- immunogenomics · VIMA software and demonstrations
Original public repository. Documentation inspected; code not executed. A permissive code licence was not independently verified.
Images & figures
Conceptual illustration, not microscopy or study data. Both upper grids contain the same mixture; their geometry differs. VIMA learns patch representations and tests associations across samples. A highlighted arrangement does not establish a disease mechanism.
Euangelion · original explanatory graphic · Source · Usage terms