Bio-Rad iQ™5 Optical System Software, Version 2.1 - Manual (Page 107)

Section 6 Data Analysis Module
- 98 -
Relative Quantity When a Control is Assigned
With a control assigned relative quantity (dC
T
) for any sample for all genes is calculated as
follows.
Where E = Efficiency of primer (primer/probe) set
this efficiency is calculated as follows (% Efficiency * 0.01 + 1)
where 100% = 2
C
T (Control)
= Average C
T
for the sample which has been assigned as a control
C
T (Sample)
= Average C
T
for the sample
This is where the calculations differ from those outlined by Dr. Jo Vandesompele on the geNorm
web site (http://medgen.ugent.be/~jvdesomp/genorm/). In the example on the geNorm Web
site, the results are not scaled the control until normalized expression is calculated. This is
referred to as rescaled normalized expression in the example spreadsheet.
Standard Deviation of Relative Quantity of Gene x for a Given Sample
SD Relative Quantity = SD C
T
sample * Relative Quantity sample * Ln(E(Gene x))
Where:
SD Relative Quantity = Standard Deviation of the Relative Quantity
SD C
T
sample = Standard Deviation of the C
T
of the sample
Relative Quantity sample = Relative Quantity of sample
E = Efficiency of primer (primer/probe) set
this efficiency is calculated as follows (% Efficiency * 0.01 + 1)
where 100% = 2
6.9.10 Gene Expression - Frequently Asked Questions
Why should I normalize my data?
Relative quantity data that is not normalized by some means is difficult to interpret. Imagine the
case where you load 1 µg of RNA in one well and 10 ng in the other well. If you perform a
relative quantity analysis on the results from such an assay the fact that the 10 ng sample has a
smaller relative quantity value is irrelevant. It is likely the result of using less RNA and not the
result of some biological response.
How does normalized expression calculated by this software compare to the value
calculated using the ddC
T
equation?
If you leave efficiencies at 100% and only evaluate one reference gene and one gene of interest
the software will generate the same results as you would get using the ddC
T
equation. The
standard deviations will be larger since the error propagation outlined in the initial publication is
inappropriate.