Analysing Localisation Microscopy data

• Analysing data not generated by PYMEAcquire
• Using an image splitting device for multi-colour ratiometric imaging

Starting analysis infrastructure

To improve performance, PYME distributes localization analysis over multiple worker processes, with a server process used for communication. These can either all be on the same machine, or be distributed across a network/cluster. The server and worker processes need to be running before starting the analysis. To launch all components on a single machine , launch PYME>PYMEClusterOfOne from the start menu (Windows), or run

PYMEClusterOfOne

in a terminal window (OSX, Linux) or “Anaconda prompt” (Windows) 1 .On Windows you may be prompted to allow python through the firewall, which you should accept.

Note

We previously used a slightly different analysis architecture, launched with the launchWorkers command. If you are familiar with launchWorkers, PYMEClusterOfOne should be a drop in replacement. The most noticeable differences will be a different task monitoring window, and that analysis results now go in an analysis subdirectory of the image directory rather than a higher level analysis directory. We’ve done a reasonable ammount of testing, but if something doesn’t work launchWorkers is still available (for now, python 2.7 only). Please also let us know so we can fix it.

Note

To distribute analysis over a computer cluster, see cluster setup.

Loading data

Once the the cluster (of one) is running, open raw blinking series with dh5view.

If the data was acquired with PYMEAcquire and saved as .h5 the localization analysis plugin should load automatically. Otherwise, select LMAnalysis from the Modules drop-down menu to activate “Localisation Mode” 3.

With the data loaded in dh5view, one should see something like:

Data not acquired using PYMEAcquire

In addition to requiring manual activation of “Localisation Mode”, data not originating from PYMEAcquire will require values for various camera parameters to be entered in the metadata (see Analysing data not generated by PYMEAcquire for details).

Ratiometric multicolor data

Special attention is also needed for analysing simultaeneous multi-colour data (see Using an image splitting device for multi-colour ratiometric imaging).

Analysis settings

The Analysis and Point Finding panes in the left hand panel control the analysis settings.

Fit types

PYME offers a large number of different fit types, and is easily extensible to support more. The most useful ones are given below. If in doubt, you usually want LatGaussFitFR.

Type	Description
LatGaussFitFR	Simple 2D gaussian fitting.
InterpFitR	Fits an interpolated PSF to localisation Data (3D)
SplitterFitFNR	Like LatGaussFitFR but for ratiometric data
SplitterFitInterpNR	Like InterpFitR but for ratiometric data (3D)
SplitterFitInterpBNR	Like SplitterFitInterpNR but for biplane ratiometric data (3D)
GaussMultiFitSR	2D Multi-emitter fitting
SplitterShiftEstFR	Used for estimating shift fields between channels
AstigGaussGPUFitFR	Fast astigmatic fitting (3D) Requires pyme-warp-drive.

Settings

Setting	Description
Threshold	The threshold for event detection. This is not a strict threshold, but rather a scaling factor applied to a local threhold computed from the estimated pixel SNR. Will generally not need to be altered (except for some of the interpolation based fits)
Type	The type of fit to perform
Interp	The type of interpolation to perform (only methods with Interp in their name)
Start at	The frame # should to start our analysis at (zero-indexed)
Background	What range of frames should be used for background estimation. Uses python slice notation.
Subtract background	Should we subtract the background before fitting (rather than just for event detection)
Debounce r	The radius (in pixels) within which 2 events cannot be reliably distinguished
Shifts	The shift field to use for chromatic shift correction (only methods with Splitter in their name.
PSF	The PSF measurement to use (only Interp methods)
Track Fiducials	Whether to estimate drift using fiduciaries
Variance Map	An image containing a pixel by pixel map of camera variance (read noise) for sCMOS correction
Dark Map	An image containing a pixel by pixel map of dark values for sCMOS correction
Flatfield Map:	An image used for flatfielding (gain correction) with sCMOS cameras

Some fit modules will also display custom settings.

Starting the fitting

Testing object detection

We use a signal-to-noise dependent threshold, which makes detection fairly robust with a threshold factor near 1 (if that the metadata is correct). It is nevertheless worth testing the detection by clicking the Test button, especially when looking at data from a new microscope or type of sample. This performs the object finding step on the current frame. If this performs poorly, one should check the Camera.ADOffset setting in the metadata (accessible through the Metadata tab) to see if this is reasonable before attempting to tweak the detection threshold. ADOffset is defined as the average dark value on the camera. We recommend imaging protocols take a number of dark frames before the turning the laser on, and will use these frames, if present, to estimate ADOffset. This estimation can be fooled if the room lights are on and/or the laser shutters are misbehaving). Incorrect camera maps (sCMOS), or read noise calibrations can also lead to poor detection. Metadata parameters can be edited by right clicking the appropriate field in the Metadata tab.

Launching the analysis tasks

Once satisfied with the event detection, the analysis proper can be started by clicking the Go button. The results will automatically be saved as .h5r files, either under the PYMEDATADIR directory (if the environment variable was set earlier), or in a directory called PYMEData in the users home directory (c:\\Users\\<username>\\ under windows).

The resulting .h5r files can be opened in VisGUI.

Localizing directly from PYMEAcquire

Localization can be initiated directly from PYMEAcquire. In the ‘Time/ Blinking Series’ panel on the right of PYMEAcquire, expand the ‘Real-time analysis’ section and click the ‘Save analysis settings to metadata’ checkbox. During acquisition, you can either click the ‘Analyse’ button in spooling progress section, or your acquisition protocol can do this automatically.

Batching analysis

To batch-run analysis of multiple series, launch the full clusterUI webapp by running PYMEClusterOfOne --clusterUI=True. This will require the optional dependency django (See also cluster setup).

Footnotes

1

Found under the “PYME” group on the start menu, if you used the installer, otherwise under the “Miniconda” or “Anaconda” group.

2

If you’re running the analysis on the data acquisition computer, and have a decent number of cores available (e.g. 8), better performance can be achieved by reserving a core for each of the Acquisition and Server processes (ie limiting the number of workers to 6 in our 8 core case). This can be done by explicitly specifying the number of workers to launch as an argument eg: launchWorkers 6.

3

Instead of manually loading the LMAnalysis module every time you launch dh5view, you can force it to start in localisation mode with the -m LM command line switch - i.e. dh5view -m LM filename

4

Using the zeroconf/MDNS protocol.