What problem does it solve?
Map and analyze spatial distributions of metabolites from MSI experiments to convert raw imzML/MSI data into interpretable ion images, segmented tissue regions, and statistically compared ROI spectra so researchers can localize biochemical signatures within tissue sections.
Core Features & Use Cases
- imzML parsing & ion image generation: extract pixelwise spectra and build ion images for target m/z values with configurable mass tolerance and normalization.
- Datacube construction & dimensionality reduction: bin spectra onto a common m/z axis to form a (rows × cols × bins) datacube and apply PCA/IncrementalPCA for scalable analysis.
- Spatial segmentation & ROI comparison: cluster pixels by spectral similarity (k-means or hierarchical), extract mean spectra per region, and run statistical tests with FDR correction to find discriminating m/z features.
- Coregistration & visualization: align MSI ion images with H&E histology and export publication-quality figures.
- Use case: Align a MALDI imzML dataset to histology, segment tumor and stroma regions, and report m/z features enriched in tumor after TIC normalization and FDR-corrected testing.
Quick Start
Load the imzML file, generate ion images for biomarkers of interest, build a datacube with 0.1 Da bins, run PCA and k-means (k=5) to segment tissue, and export ROI comparison results.
