Embedded Files
1. Overview
The Nanotime Display page, which is available with smFRET (TCSPC), smFRET ns-ALEX and Polarization Anisotropy Measurement Types, has several purposes (note 1, 2).
First, it represents the histograms of photon arrival times with respect to a synchronization signal (generally, the laser pulse following the detected photon) in the Nanotime Histograms graph. There is one histogram plot per channel (Donor and Acceptor).
Second, these histograms can be used to define the donor and acceptor emission periods (for ns-ALEX data), as well as the time windows used to compute average nanotimes for each photon stream within each burst (note 3).
Third, the page comprises a second nanotime histogram display (Constrained Bursts Nanotime Histograms) where sub-population nanotimes are histogrammed. Some derived quantities are computed and displayed next to this second graph (see Section 6).
Finally, these histograms can be used to select different decay curves to be analyzed in the Nanotime Histogram Analysis window in combination with display in the Nanotime Analysis page.
The top graph (Nanotime Histograms) is generated when pressing the Update Nanotime Histograms button. The graph contains two plots (in general) corresponding to the histograms of photon nanotimes in the file (Donor and Acceptor channels), converted to ns units using the TAC specifications provided with the file (visible in the SPC Parameters control on the Data Page).
If a valid Nanotime Histogram Calibration File Path has been provided on the Data Page, the file is used to correct the nanotime histogram for any non-linearity of the TAC. Such a calibration file can actually be generated using the tools provided on the Nanotime Display page as discussed in the last section of this manual page.
2. Defining the ns-ALEX emission windows
If the ns-ALEX emission windows are not specified in the file (currently, only photon-HDF5 files contain such a definition), or if the user chooses to ignore these file definitions (by checking the Ignore File ns-ALEX Period Definitions checkbox), they need to be defined with the help of cursors as shown in the figure below. Note that the histograms do not need to be plotted to define ns-ALEX periods, but it clearly helps.
Unless this has already been done previously, it is necessary to specify a few control parameters affecting the interpretation of the raw data as well as the histograms calculation, as discussed next.
2.1. Binning X
The Binning X parameter and the associated Effective Bin (ns) indicator (red box in the Figure above) allows adjusting the nanotime histogram bin in multiple of the TAC (or TDC) resolution. Changing this value automatically recomputes the histogram.
Multispot: In some cases, multispot data might be acquired with TACs (or TDCs) having different calibration (bin size) for each channel. In this case, the value displayed in the Effective Bin (ns) indicator will reflect the corresponding size, which can be found in the ALiX Settings >> Lifetime Histograms tab, TAC Bin Sizes (ns) array. This array is automatically filled with information saved in the data file (if present). If no information was found in the file, the average TAC bin will be used. The user can enter channel-specific bin sizes in the array just mentioned.
2.2. Acceptor Offset
In general, the time zero of a TCSPC histogram is not trivial to define because of several considerations related to cabling, electronics, optics, etc. Ideally, the Donor and Acceptor channel histograms should show the laser pulse located at the same position in time. In practice, there can be a slight offset (<100 ps) between both. The Acceptor Offset (ns) parameter (blue box in the Figure above) allows shifting the Acceptor channel histogram with respect to the Donor channel histogram so that both are aligned in time.
Changing this parameter automatically shifts the Acceptor histogram.
A convenient way to find the optimal shift to apply is by using an instrument response function (IRF) file consisting of a sample emitting photons with as little a delay as possible with respect to the laser excitation pulses, preferably in both channels (see Figure below for an example).
With such a file, checking the Autocorrect Channel Shift checkbox will attempt to align both histograms using one of two possible algorithms selected in the Channel Offset Correction Algorithm pull-down list:
CCF: finds the offset maximizing the cross-correlation function between the two histograms
Peak Location: finds the peaks of each histogram and aligns them
If successful, the Acceptor Offset (ns) value is automatically updated and a Guessed Period (ns) (if found) is displayed instead of the user-entered Period (ns) parameter (see below). This value can be overwritten by simply typing in the correct value (which will then revert to the Period (ns) designation).
Multispot: In the case of multispot data, each TAC (or TDC) may define time zero in a different manner, in which case an histogram shift needs to be applied for each channel. These parameters are defined in the Lifetime Decay Shifts (ns) array in the ALiX Settings >> Lifetime Histograms tab. This array is auto-filled with information obtained from the data file (if found) or can be entered by the user.
Note that the histogram shift is different from the Acceptor Offset (ns) and adds up to it.
2.3. Period & Number of Periods
The Period (ns) and # Periods parameters are used in the Nanotime Analysis page to enforce periodicity of the nanotime histograms.
The Period (ns) parameter normally corresponds to the exact laser period used in the experiment. However, it is possible to artificially decrease or increase it for numerical purposes. For instance, if the period is 50 ns but only 40 ns worth of histogram are usable, then it might be preferable to discard part of the histograms by enforcing a shorter period.
The # Periods parameter is in general equal to 1 but can be larger if the TAC (or TDC) time window encompasses more than one laser period. In this case, the Donor and Acceptor windows (see next section) should be defined once and are assumed to be repeated identically with a shift of one, two, etc. laser period(s).
2.4. Donor & Acceptor Emission Windows Definition
As for us-ALEX data, ns-ALEX data needs to be separated into Donor and Acceptor emission windows in order to allow further analysis of the data. This definition is done using nanotimes rather than macrotimes, due to the typical laser repetition frequencies used in fluorescence decay analysis.
To define the Donor and Acceptor windows, use the (D Start, D stop) and (A Start, A Stop) cursors (green box in the top Figure). There are several predefined relative positions for these windows, selectable from the ns-ALEX Cursor Mode pull-down list. In general, Free Boundaries will be used, in order to allow for some gaps between emission periods.
Note that in the case where multiple laser periods fit within one TAC window, it might be necessary to specify several sub-windows for the Donor and Acceptor emission. In this case, only the first set of Donor and Acceptor windows needs to be defined, and a # Periods value larger than 1 specified. The other Donor and Acceptor emission windows are simply offset by one laser period from the previous windows.
To validate this definitions (and thus set the stream to which each photocount in the data file belongs), press the Define ns-ALEX Emission Periods button (top Periods button).
To later reset the cursors to their position corresponding to this definition, press the Restore ns-ALEX Period Definitions button next to it.
3. Defining the average burst nanotime calculation windows & Computing the average burst nanotimes
The previous definitions are in general not appropriate to calculate time-to-laser time lags (also referred to as photon nanotimes) because they tend to start before the laser pulse (this is useful for decay fitting) and end up well before the other laser excitation pulse.
A better definition to compute nanotimes requires setting the beginning of an emission period to the peak of the IRF, and its end as close as possible to the peak of the IRF of the other laser. Nanotimes falling outside these boundaries are set to NaN (not a number), which means that they are not used in calculations involving nanotimes.
Moving the cursors to these locations and pressing the Compute Photon Nanotimes button uses these definitions, and updates the individual photon nanotime accordingly (photon nanotimes are in general undefined when a file is loaded; only their nanotime counter value - in units of TAC resolution - is provided): their corrected nanotime is stored as well as the emission window
Furthermore, if bursts have been defined (by a burst search), the average nanotimes within each burst will be computed and stored.
Note that there are 4 average nanotimes for each bursts, corresponding to the following photon streams: F_D^D, F_D^A, F_A^D, F_A^A. If bursts have not yet been defined, the individual photon nanotimes will be used to compute these average burst nanotimes next time a burst search is performed.
Since the same cursors are used to define Emission Periods and Nanotimes, the Restore Nanotime Cursor Definitions button next to the Nanotimes button can be used to bring them back to the Nanotimes calculation windows boundaries, if they have been moved.
In other words, the two Restore... buttons allow switching back and forth between definitions.
When the calibrated photon nanotimes have been defined, the 3rd LED of the Task Completion Indicator at the bottom right of the page lights up.
For data files which have no nanotimes associated with photons, this LED remains dark.
4. Displaying and analyzing subpopulation nanotime histograms
When bursts have been selected in the Burst Analysis or in the ALEX Analysis page, subpopulation nanotime histograms can be computed and displayed in the bottom graph (Subpopulation Nanotime Histograms graph) by checking the corresponding checkboxes (ALEX-selected Bursts and/or Constrained Bursts) and pressing the Selective Nanotime Histogram button (Histogram) to the right of the graph's legend.
As an example, the Figure below shows the D-only population nanotime histogram corresponding to the same data represented in the Figure at the top of this page:
The donor decay (green curve, right side) is clearly mono-exponential as expected, in contrast to the corresponding decay shown in the top Figure, which encompasses all donor channel photons detected during the experiment (including doubly-labeled molecules, but also inter-burst background signal).
A similar analysis could be performed for the A-only or FRET sub-populations, or for a subset of bursts selected with a particular set of selection criteria applied in the Burst Analysis page.
The Outside Bursts checkbox allows plotting the nanotime histograms of photons found outside the selected bursts:
In this particular file, the majority of photons is found outside bursts, and the interburst decay histograms are very similar to the whole time trace histograms shown in the top Figure.
5. Selecting histograms for nanotime analysis
The right-click menu in the plot region of both graph gives access to a few options:
Choosing any of the Use as... item selects the plot closest to the current cursor location and stores it as the corresponding curve in the Fluorescence Decay Curves graph of the Nanotime Analysis page (see the corresponding manual page for details).
There are 3 different possible sub-populations: D-only, A-only and FRET, to which two decays can be associated: Donor and Acceptor.
IRF plots
In general, the IRF plots will be chosen in the top graph, as the corresponding data files are generally ensemble measurements (there is no need to perform a burst search or burst selection prior to any kind of analysis). The IRF plots are treated in a specific manner when exported to the Fluorescence Decay Curves graph of the Nanotime Analysis page:
the baseline (calculated as the location of the count histogram maximum) is subtracted from the nanotime histogram.
the resulting plot is normalized such that the integral of the remaining curve is equal to 1.
IRF plots are associated with the right-hand side scale (IRF Amplitude) of the Fluorescence Decay Curves graph.
Only the part of the plot within the min and max cursor positions is retained (and used in the calculation detailed above).
Fluorescence Decay Plots
As for the IRF, only the part of the plot within the min and max cursor positions is retained. The main difference is that:
the histogram is exported unchanged to the Fluorescence Decay Curves graph,
the histogram can be a burst-selected decay from the Subpopulation Nanotime Histogram graph,
the plot is associated to the left-hand side scale (Amplitude).
Multispot data
If the file is a multispot data file, an additional option is available for nanotime analysis. By right-clicking on the plot of interest while pressing the Ctrl key, the following shortcut menu appears:
Instead of replacing the corresponding decay curve, the plot is added to the existing decay curve (if there is no decay curve in memory, the selected plot becomes the decay curve), allowing to pool data from multiple spot into a single decay curve. For this to make any sense, it is important to make sure that all plots have the same time zero location and that their IRFs are similar.
6. Nanotime correction parameters
To be added
7. Saving (and Loading) a nanotime histogram calibration file
7.1. Saving
TAC (or TDC) might exhibit some non-linearities visible in the nanotime histogram of an uncorrelated sample as ripples in the baseline.
The best way to acquire data for nanotime histogram calibration is to illuminate the detector with a constant signal (e.g. attenuated room light) while still triggering acquisition with the pulsed laser which will be used in the actual experiments. The resulting histogram should be for the most part flat, with some residual oscillations. Obviously, all acquisition parameters (including TCSPC electronics settings) should be identical to those used during the actual experiment.
The data file should contain enough photons such that the baseline noise level is minimal (ideally, less than 1%), since the calibration process consists in dividing the raw nanotime histogram by the calibration histogram. Any noise in the latter will be injected in the former. In general, this results in rather large photon-counting data files, therefore it is useful to:
either export these histograms directly from the acquisition software (currently, only the Becker & Hickl SPCM software .std format is supported)
or load the file in ALiX, plot the nanotime histograms nad save the resulting histogram (or set of histograms if the data comprises two channels) as a calibration histogram file (.spch format, specific to ALiX).
The second option requires loading the file from the Data File page, selecting .spc as the Calibration File Type and using the Nanotime Histogram Calibration File Path control to choose the file.
This action triggers a series of action resulting in the generation of baseline-corrected and normalized calibration histograms ready for saving. Saving those histograms as .spch format is performed by pressing the Save Calibration Histograms button (Save Cal. Histograms).
7.2. Loading
There are 3 different ways to load a calibration histogram using the Nanotime Histogram Calibration File Path control on the Data File page, depending on the Calibration File Type value.
.spc File: see the explanations in Section 7.1. Note that this is usually a time and RAM-consuming process.
.sdt File: an histogram file needs to have been exported from the Becker & Hickl SPCM acquisition software.
.spch File: a calibration histogram file exported from an original .spc photon-couting data file as explained above.
If there was a ns-ALEX data file (or compatible file) loaded before this action, their incompatibility will be indicated by a LED next to the Nanotime Histogram Calibration File Path control. In case of incompatibility, the calibration file will not be loaded in order to not interfere with possible later analysis using the original file.
Therefore, in order to apply a nanotime calibration file to a a data file for the first time, it is necessary to first load that data file, then load the calibration file (and verify that they are indeed compatible), and finally, reload the original data file in order for the calibration to be applied. Any later file compatible with the currently loaded calibration can be loaded directly. The calibration will be applied automatically.
Notes:
The tools on this page behave slightly differently in the case of multispot data files. Check the corresponding Multispot sub-sections for details.
Some of the definitions (such as Donor and Acceptor emission windows) and actions (such as defining the previous windows) discussed here are irrelevant for single-laser excitation data (e.g. polarization anisotropy files).
The windows defined in the Nanotime Dispay page are, strictly speaking, excitation windows, rather than emission windows. Indeed, all we can tell with certainty is after what laser pulse each photon is detected. While it is in general likely that this laser pulse is responsible for the excitation which resulted in the emission of that photon, this is not necessary the case. For instance, in the D-only nanotime histograms shown above, the donor channel decay is not completely over at the end of the Donor window, and some photons will be detected in the next Acceptor window. Those photons should in principle be labeled F_D^D but will in practice be labeled as F_A^D. This is the reason why special correction procedures and coefficients are needed in ns-ALEX analysis of photon-counting data. With these caveats in mind, the term emission window will be used here.
Page updated
Report abuse