Data analysis in ALiX proceeds as indicated in the workflow below:
1. File
The first step consists in loading a file in memory.
Due to the variety of file formats and structures, loading a file is usually a three step process, with a first "preview" step allowing basic information on the file content to be presented to the user, followed by a second step involving choosing which channel will be loaded and which channel it will be designated as (donor, acceptor, monitor, or other types depending on the file type). The last step consists in loading the selected data.
2. Alternation Period Definition
For data having been acquired in a dual-laser, alternated excitation setup, definition of the alternation period is required. Alternation period definition is done in either of two pages, depending on the file type.
Data sets acquired with a single laser do not require this step and therefore the corresponding page may be missing.
For us-alternation, the us Alternation Periods Page will be available, while for ns-alternation, the Nanotime Display Page will be available.
3. Time Traces
Time Traces can observed as soon as the different streams are identified. However, no burst will be identifiable and no background correction will be applicable. For this and other features to be available, such as burst statistics time traces, further analysis steps need to be completed first.
4. Count Rate
As for time traces, count rate histograms can be computed right after the different streams have been defined. Burst count rate calculation will not be operational until after a burst search has been performed.
5. Background Rates Calculation
Background rates are necessary to perform a burst search and compute background-corrected quantities. They are also necessary for other background-dependent calculation in other pages. This is therefore a crucial step of the analysis that can hardly be skipped.
6. Time-dependent Background Rates Calculation
An advanced method to compute background rates consists in defining them over short time windows. This is sometimes necessary if the background rates in the sample change over time due to evaporation or adsorption of molecules. It is not a mandatory step and can be skipped.
7. Burst Search
Burst analysis requires the definition of bursts. This is performed on this page. It is arguably the most important step in the analysis.
8. Burst Analysis
After bursts have been defined, a variety of analysis can be applied to them. At the very least, further selection among the bursts founds during the search can be done by applying constraints to various burst statistics. This is the first role of this page: defining constraints to be applied to select bursts for further analysis.
Analysis can consist in histogramming various statistics, or plotting one statistic versus another, calculating burst rates (the number of burst per unit time), etc, all of which can be done in secondary pages within the the Burst Analysis page.
It can also be performed in other tabs, whenever a "Use Burst Analysis Constraints" checkbox is present as for instance in ALEX Analysis, Time Trace or Correlation.
9. ALEX Analysis
ALEX analysis consists in exploiting the E-S histogram (also known as ALEX histogram) to identify sub-populations within a data set.
The ALEX histogram can be constructed using all burst found during the burst search, or limited to bursts verifying the constraints defined (and applied) on the Burst Analysis page.
Projection histograms (E and S histograms) can be calculated and fitted with a set of pre-defined functions.
Once a region of interest (ROI) has been defined in the the ALEX histogram, analysis on several pages of ALiX can be limited to the set of bursts within the ROI.
Â
10. Gamma Analysis
Gamma analysis is a multistep process and consists in establishing a relation between 1/S and E for a few samples (a minimum of 2). The result of fitting a linear relationship between these two observables is the value of the gamma factor.
11. Correlation
Autocorrelation and cross-correlation analysis can be perforrned on different streams and limited or not to constrained of ALEX-selected bursts. If done on the whole dataset, this action can be performed immediately after the different photon streams have been defined.
12. Shot Noise Analysis
Shot noise analysis (aka probability distribution analysis) analyzes the proximity ratio (or E) histogram in terms of the burst size and duration distributions and a model distribution for either the FRET efficiency or the distance between dyes, trying to optimize the latter in order to obtain a theoretical histogram as close as possible to the observed one.
13. Scripting
A number of repetitive tasks can be automated with the help of a few configurable scripts available on different subpages of the Scripting page. In the particular case of multispot data files, several scripting tools simplify the task of performing the same type of analysis on different pairs of channels or allow pooling the data from multiple spots before analysis.