I am interested in opinions from this esteemed group concerning the best way to analyze the following data (see attachment in MSWord format). My colleagues and I can not reach a consensus. We are assessing cell surface determinants on eosinophils in lysed whole blood. Since lymphocytes are also analyzed, compensation is set based on FITC-CD3/PE-CD19. An eosinophil gate is established by back-gating on CD45bright/CD14 dim (autofluorescent) cells and a CD16/CD14 cocktail is used to exclude any neutrophils or monocytes that might contaminate the gate. To determine the % positive cells, we are using either the dot plot or histogram. For dot plot, a tight gate (R2) is set around autofluorescent cells in the PE-IgG control sample (Top left panel). A larger gate (R3) is established to allow identification of positive cells. The % positive cells are calculated as (1-(R2/R3))X100. Histograms are gated on R1 (eosinophils on scatter plot) and R3. M1 is set on isotype control. For signals that have a large shift (ie. CD11b, middle panel), the % positive are the same using either method. When the signal is weak, such as for cytokine receptors (right panel), only a small proportion (in this case 27%) are positive based on histogram; whereas, the dot plot would indicate nearly 100% positive based on a shift out of the "negative" region. Any opinions, comments about how to express these data? Is there any reason why the dot plot method would not be acceptable?
This archive was generated by hypermail 2b29 : Sat Mar 10 2001 - 19:31:21 EST