Quantify microscopy images and gels

What this involves

Quantifying biological images requires raw intensity arrays rather than display exports like JPEG files. Whether you are analysing western blots, colony assays, or multi-channel micrographs, extracting valid numbers demands that you remove background gradients, separate touching boundaries, and establish that your signal falls within the linear response range of your imaging detector.

For microscopy, the platform uses scikit-image, OpenCV, tifffile, and Cellpose to run background subtraction, adaptive thresholding, distance-transform watershed segmentation, and feature extraction. It measures colocalisation statistics against controls, tracks particle motility with trackpy, and calculates baseline-normalised transients in calcium imaging. For multi-channel whole-slide tissue images or imaging mass cytometry, you can execute specialised pipelines such as nf-core/mcmicro and nf-core/imcyto.

For blot densitometry, valid results depend on raw band integration, background correction, and normalisation against total-protein stains or verified loading controls across independent biological replicates. Every calculation runs on the platform's execution environment, outputting measurements, parameter uncertainties, and residual plots directly to your workspace.

How it goes

1. 1Background correctionNon-uniform illumination and baseline drift are subtracted across image channels using rolling ball or median background filtering in scikit-image.
2. 2Object segmentationBoundaries are delineated using Otsu thresholding, distance-transform watershed in OpenCV, or CPU-based Cellpose models, followed by visual inspection of overlaid masks.
3. 3Signal and spatial quantificationMorphological features and intensities are extracted with skimage.measure.regionprops_table, while particle tracks and mean squared displacements are computed using trackpy.
4. 4Colocalisation and normalisationThe agent evaluates Pearson's r and Manders' M1/M2 coefficients against rotated controls, or integrates blot bands normalised to loading controls across independent replicates.

Start it here

Ask in plain language. The agent works on a Linux machine of its own — it installs and runs what the job needs, keeps every file, and reads the results back with the numbers explained. Nothing to set up, and a long run continues after you close the page.

What to watch for

• Saturated bands or blown-out cores have flat intensity profiles with no quantitative meaning and understate the strongest lanes.
• Quantification performed on contrast-adjusted exports or lossy JPEG files distorts intensity relationships and invalidates densitometry.
• Apparent yellow pixels in merged channels do not establish colocalisation; compute Pearson's or Manders' metrics and rule out bleed-through and chromatic shift first.
• GAPDH and actin saturate readily and make poor loading controls opposite strong signals; prefer total-protein stains.
• Technical replicates from the same biological sample must be averaged into a single n rather than treated as independent measurements.

What it runs

Questions