← back to the tutorial

Finding your way around SMAPpy

The words of this tutorial, step by step. The text lives in src/smappy/tutorial/topics/; the directions it follows are in src/smappy/tutorial/STYLE.md.

Two windows

1. card: Finding your way around SMAPpy

The two windows, the tabs, and the four ideas the rest of SMAPpy is built on: filters, layers, grouping and the selection.

Everything here uses simulated data, so you can follow along: File → Simulate.

spoken: Welcome. This tutorial is a tour of SMAPpy: where things are, and four ideas that everything else is built on.

2. SMAPpy opens two windows. On the left, the control window, with every setting and tool.

3. On the right, the render window: the super-resolution image itself.

Getting data in

4. Open a file from the File menu, or with Ctrl+O. SMAPpy reads localizations from the common fitting programs.

5. Add file... puts a second file into the same table, for instance the next field of view.

6. The File tab has the same, and more: saving, exporting an image, and simulating data.

7. To practise, Simulate makes a dataset with a known structure: a ring and two crossing lines.

8. Click Run.

9. A moment later the localizations are in, and the picture is drawn.

The tabs

10. The tabs follow the workflow, from loading and fitting to the picture and its analysis.

Looking around

11. In the render window, scroll to zoom and drag to move.

12. Zoomed in, the ring is made of single localizations.

13. The overview in the Render tab shows the whole field of view. The yellow box is where you are; click to move it.

14. Reset view, or Ctrl+0, shows everything again.

15. The View menu has Reset view too, next to the 3D view.

Filters

16. card: Filters

Not every localization is a good one: dim ones are imprecise, and a bad fit lands in the wrong place.

A filter keeps the localizations whose values lie in a range – precision below 10 nm, say. It changes what is drawn and what the analysis sees, never the data: clear it and everything is back.

spoken: Filters. A filter keeps the localizations whose values lie in a range. It changes what is drawn and analysed, never the data.

17. A quick button chooses what to filter on. The shaded part of the histogram is the range kept: drag its edges, or type min and max.

18. With a tighter maximum, fewer localizations pass: the count is above the histogram. The rendering tutorial has more on filters.

Display

19. Display sets how the layer is drawn: render mode, colour table, contrast.

20. Colour by a field, here z, and the picture becomes a depth map.

Layers

21. card: Layers

A layer is one way of drawing the same localizations: its own filter, colour table and contrast. The visible layers are added into one picture.

Two colour channels are two layers. So are the first and the second half of an acquisition – which is how to spot drift at a glance.

spoken: Layers. A layer is one way of drawing the same localizations, with its own filter and colours. Visible layers add up.

22. The layer strip: a button per layer, + to add one. Everything below edits the highlighted layer.

23. Layer 1 now keeps the first half of the frames, drawn in red.

24. Layer 2 keeps the second half, in green.

25. Where the halves agree, the picture is yellow. Red and green fringes would mean drift.

Grouping

26. card: Grouping

A fluorophore is often on for several frames, and is localized in each of them.

Grouping links localizations that are close (50 nm) in consecutive frames (one dark frame allowed) into one, with the photons added. The ungrouped table stays; each layer chooses which one it shows.

spoken: Grouping. One blink seen in several frames is several localizations. Grouping links them into one.

27. SMAPpy groups a new table straight away, so what you see are blinks, not single frames. More on grouping in the rendering tutorial.

The selection

28. card: What a plugin works on

Every analysis gets the selection: the localizations of the current layer that pass its filter and lie inside the region of interest, if one is drawn.

What you see is what is measured.

spoken: The selection. A plugin measures what passes the current layer's filter, inside the ROI if one is drawn.

29. To draw a region of interest, click ROI. Right-click it for a line, rectangle or polygon.

30. Then drag over the picture. The toolbar shows how many localizations are inside, and plugins now measure only those.

Plugins

31. The Analysis tab has a section per plugin, one open at a time.

32. Find narrows the list; + pins another plugin to the tab.

33. Localization Statistics measures photons, precision and on-time.

34. Run it: a summary appears, and Plot draws the figures.

35. The figures get a window of their own, reused by the next run.

36. The Plugins menu has every plugin, to open once in a window of its own. Find a plugin..., Ctrl+Shift+P, searches them by name.

Undo and saving

37. What changes the table, like a drift correction, can be undone. Grey here: filters and layers never change it.

38. Save, or Ctrl+S, keeps the localizations, their history and your ROIs in one file.

Next

39. card: That is the tour

- Filters choose which localizations count.
- Layers draw the same localizations in different ways, and add up.
- Grouping turns one blink seen in several frames into one localization.
- The selection – filter and ROI – is what a plugin measures.Next: fitting raw data, rendering in 2D and 3D, and the ROI manager.

spoken: That is the tour. Next: fitting raw data, rendering, and the ROI manager.