Skip to content

Spectrum library and decomposition

CytoSpec works with two kinds of reference spectra:

  • the dye library shown by the Spectra window - published absorption and emission curves, for reference;
  • the spectrum database - spectra measured with your own instrument and saved from the Average window, used for decomposition.

The Spectra window

Tools → Spectra opens a list of dyes. Type into the search box to narrow the list; click a dye to plot its absorption and fluorescence curves.

Tick Multi to add curves to the chart instead of replacing them. Clear Chart removes all curves.

The library lives in the spectra folder next to CytoSpec.exe: a names file listing the dyes - each dye name on one line, its file name on the next - and one file per dye with wavelength/absorption and wavelength/fluorescence pairs, comma-separated, after a header line.

Building the spectrum database

  1. Open or gate a data set that contains only one dye - a single-stained control.
  2. Open Analyze → Average.
  3. Press Save as Spectrum, give the spectrum a name, and tick Normalize to scale it to a maximum of 1.

The spectrum is added to the database and is available in the Decomposition dialog.

Decomposition

Tools → Decomposition expresses the spectrum of every event as a combination of reference spectra - spectral unmixing.

  1. Select the data set and choose the command.
  2. In the dialog, tick the channels to use. Leave out channels that carry no signal or only noise.
  3. Tick the spectra of the dyes present - two or more. Click a spectrum to preview it.
  4. Press OK.

A new data set is created with one channel per chosen spectrum, named after it, holding each event's amount of that dye. Plot it like any other data set.

Delete checked removes the ticked spectra from the database.

Warning

The spectra must have been measured with the same channels as the data. If a spectrum is shorter than the highest channel chosen, the decomposition is refused.

Finding signal events

Tools → Find Signal shows the distribution of each event's trigger value and asks for a threshold. The events above it become a new data set - a quick way to separate real events from background in raw detector logs.