A checklist of important parameters in iSMS

This page is meant to provide smFRET beginners and new users of iSMS with a checklist of parameters to pay particular attention to when analysing data from your setup for the first time. The list is by no means fully comprehensive - more information may be found in the documentation pages and by playing with the demo data sets provided with the software. But make sure you are familiar with concepts and parameters below before you interpret your data in detail.

1. Excitation scheme and order

Specify the excitation scheme of your data: Single-color (regular) or two-color (ALEX/PIE). Single-color data is recorded using a single laser excitation wavelength source (excitation of donor only). Two-color data is recorded using alternating excitation with two different laser excitation wavelengths (one for the donor and one for the acceptor).

In two-color excitation, iSMS automatically registers the excitation order of the raw data, i.e. which excitation wavelength corresponds to which frame in the raw video. This prediction is accomplished using the default positions of the two emission channel ROIs. Thus, in ALEX-mode make sure the default ROI positions correspond to, or at least overlap, the relevant emission channel regions of the image. If the two ROIs are not positioned properly by default then set the position of the ROIs and save the new ROI positions as the default (from the File-menu) and reload the data.

See: Excitation schemes.

2. Camera background and offset

OBS: If you are in doubt about your camera background and offset value you are able to use the pixel inspection tool from the Exploration menu to check the pixel values of your raw data. See this page for more information.

Often a constant offset value is applied by the camera acquisition software to all pixels in the movie (e.g. 100 counts) in order to avoid negative pixel values (to save memory). If not taken into account, the camera background and offset will have an effect on the magnitude of the raw background counts calculated by iSMS. However, as long as the specified offset value does not exceed the actual background count of your raw images the value of the offset will not influence the background-corrected intensity traces nor the FRET efficiency.

See: Camera background.

3. Molecule density

We recommend to have a density corresponding to 20-100 molecules in a 512x256 frame. More dense samples will challenge the automated emission channel alignment, the identification of donor-acceptor pairs and the estimation of local backgrounds. In addition, the point spread function of each molecule is often broader than what most people think - use the contrast sliders above the images in the main window to visually inspect how much the signal from your neighboring molecules overlap.

4. Emission channel alignment

Make sure the two emission channel ROIs are aligned properly every time you run the peak finder.

See: Align emission channels.

5. Peak finder settings

Note that the peak finder runs on the average image interval defined by the frame slider above the image frame in the main window. Use the contrast sliders above the images in the main window to guide your eye on optimal intensity thresholds. It is not critical whether the peak finder returns the exact number of molecules in the data as false positives can be removed during subsequent analysis (using the FRET-pairs trace window). The resulting molecules can be checked for single step bleaching and crowding and false positive removed using the FRET-pairs trace window.

See: Find FRET-pairs.

6. Donor-acceptor distance threshold

Set a proper value for the maximally allowed distance between donor-acceptor fluorescence spots defining a FRET-pair within a CCD image. The default threshold distance is 5 pixels in between the donor and acceptor spots (in the aligned CCD image frame). For very dense samples you may wish to lower this value while for sparse samples you can increase this threshold to include more FRET pairs near the edges of the ROI image where chromatic abberations can be present.

See: Find FRET-pairs.

7. Integration and background settings

As a rule of thumb, set the molecule integration aperture width (in pixels) to be no larger than 2-3 times the FWHM of your molecule point spread function and set the aperture size to be equal for the donor and acceptor. You can use the tools provided in the Settings->Integration settings dialog to estimate the FWHM of your molecules. The default aperture width is 5 pixels which is optimized for molecules with a PSF FWHM of ~2 pixels and the default background aperture. The larger the integration aperture the lower the signal-to-noise ratio you will get.

We recommend a background ring aperture with a spacer of at least 2 pixels outside the molecule integration aperture. This background estimator is generally robust provided that the background aperture is located at a radius >3 times the FWHM of your PSF. The background ring aperture can be considerably perturbed by the presence of neighboring molecules. In this case, we recommend using the intensity after bleaching as a measure of the background - but only when you are sure that bleaching is properly defined for all molecules.

See: Intensities and Background.

8. Correction factors

The correction factors (gamma, donor leakage and direct acceptor) have a large influence on the value of the FRET efficiency. If there is a high degree of uncertainty about your correction factors we strongly recommend reporting the Proximity Ratio (PR) instead of the FRET efficiency. The PR is the "FRET" calculated by setting the following correction factor values:

Proximity Ratio: gamma = 1; donor leakage = 0; direct acceptor = 0

See: Correction factors.

9. Frames and molecules included in histograms

The FRET histogram can either include several frames from each molecules or one average value of each molecule. When plotting a multi-frame histogram (the default) long-lasting molecules may contribute more frames to the histogram than molecules displaying fast bleaching. To avoid bias towards a few number of long-lasting molecules we recommend setting a max. number of around 50-200 frames included from each molecule in the histogram depending on the time-scales in your experiment.

Finally, the particular selection of molecules being plotted in the FRET histogram window can naturally have a high influence on the appearance of the histogram. Make sure you know which molecules and frames are actually being plotted and report the total number of molecules included in the histogram (you can see this value in the FRET histogram window).