Q B and S Calculator

This calculator gives a more informative view of the sensitivity of the Flow cytometer. It calculates the Quantum efficiency (Q) and the Background noise (B) of the cytometer [1,2,3]. From these two values the Predicted S or ‘separation parameter’ of the instrument is calculated. It also reports the S result from the Resolution calculator.

The Q or quantum efficiency of the cytometer, is a measure of the instruments ability to collect fluorescent photons and convert them into photoelectrons. Q is measured in molecules of equivalent soluable fluorochrome (MESF^-1).

B represents the amount of background noise in the system and is reported in MESF units [1,2].

Predicted S is calculated from the Q and B using the Wood[3] method, while B is calculated using the minimal bead-set method [1].

Data Collection

Analyse a mix of 8-peak Rainbow beads and DULL (FITC) beads with a known MESF value (used in the conversion of channel numbers into MESF units). Record the mean (Geometric means if LOG is used) or median (Linear), SD and CV from the Rainbow Blank and 3 dullest peaks. Record the CV of the Rainbow brightest peak and the mean (geomean or median) and MESF value of the DULL FITC+ reference beads.

Results

Q give the total number of photoelectrons produced per equivalent fluorochrome. Values range from 0.001 to 0.05, meaning if Q is 0.001 then we need around 1000 equivalent fluorochromes to generate one photoelectron!

B measures the noise in the system and is given in MESF values.

S or separation parameter describes the degree of resolution of two peaks, the bigger the number then the better the resolution. S is the version as used by the Resolution Calculator.

Predicted S is calculated from the Q and B values and is a theoretical resolution, again the higher the number the better.

R² (R squared) R2 (Root mean square or R squared) is a check on the quality of the linear regression of the three dullest beads and of the calculated Q value derived from it. If R is less than 0.97, then there could be something wrong in the acquisition of the data.