Skip to content

Repository files navigation

SquidXplorer

A local viewer for finished Squid HCS acquisitions. Open a plate, explore any well or region in its own napari window, and run your processing operators on exactly the wells or ROIs you pick. Read only: it never changes your acquisition and never runs the microscope.

The idea

The plate is the root. Selecting wells opens an independent napari window over them; drawing an ROI inside a window opens a child window over that region. Every window gets an integer id, and they are all collected in the Window navigator on the left.

SquidXplorer layout: the wellplate root with the Window navigator and bulk Operators, each selection opening a window with its own 2D/3D and Operators panel over a napari mosaic, and ROI child windows.

Layout (from the design deck):

  • Root: the Window navigator (a selectable list of open views) and the bulk Operators, above the Wellplate view with its Selection.
  • Each window: a 2D / 3D control and Operators for that window, over the napari mosaic of the full well, with ROI boxes you can send to 2D or 3D, and a region slider <> A1, B6, C3 ....
  • An ROI child window is the same, with a slider over its ROIs <> ROI1, ROI2, ROI3 ....

Notes:

  • Each selection from the wellplate opens a new region view. Each window gets a positive id, collected in the Window navigator on the left.
  • Windows not currently being manipulated halt their draw and refresh.
  • A memory bar warns you before the system runs low.

Operators

Processing runs your own tested implementations, called directly, never a reimplementation:

Operator Backend
Deconvolution (Richardson-Lucy, vectorial PSF) petakit
Stitch and flat-field tilefusion
Background subtraction bgsub
Nuclei detection Cellpose
Maximum intensity projection built in

Output is byte-identical to the standalone repos, pinned by tests/test_operator_fidelity.py. Results are OME-Zarr layers you toggle on and off; the raw data on disk is never touched.

Setup (Windows, one time)

  • You need Python 3.10, 3.11, or 3.12. If you do not have it, install from https://www.python.org/downloads/ and tick "Add python.exe to PATH".
  • Open PowerShell in the tool folder and run: powershell -ExecutionPolicy Bypass -File scripts\Setup-Windows.ps1
  • This puts a SquidXplorer shortcut on your Desktop.
  • To update later: git pull in the folder, then open the icon again.

Open an acquisition

  • Launch SquidXplorer. A small console opens beside it; that is normal and shows progress.
  • File, then Open acquisition folder, and pick the acquisition (the folder holding the 0 folder and/or the ome_tiff folder).
  • It reads both Squid formats (individual TIFFs and OME-TIFF), on 384 and 1536 plates.
  • The first open downsamples every well to build the plate view; opening the same acquisition again is a cache read (measured: 15.2 s to 0.08 s on a 1536-well plate). Those thumbnails go in your own cache folder, never into the acquisition: ~/Library/Caches/squidxplorer on macOS, %LOCALAPPDATA%\cephla\squidxplorer\Cache on Windows, ~/.cache/squidxplorer on Linux. Delete that folder any time to reclaim the space (one 91 MB file per 1536-well plate); it rebuilds itself. SQUIDXPLORER_CACHE_DIR moves it, and SQUIDXPLORER_PLATE_CACHE=0 turns it off.

Explore

  • Click a well to select it, Shift-drag a box or Shift/Ctrl-click to select several, then Open view to open them as one window.
  • Inside a window: 2D / 3D, draw an ROI and send it to its own child window, and run an operator on that view.
  • The Window navigator lists every open view; select rows to highlight their wells on the plate, and Collapse all when the desktop gets busy.

About

No description, website, or topics provided.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages