Skip to content

Calculation of normalised red and green intensities

2 messages · Binita Dutta, Henrik Bengtsson

Binita Dutta
# 
Dear Sir/Madam,

I could succesfully normalise my microarray data using marrayNorm package. 
However, i have not been able to get normalised red and green channel 
intensities through R package.  Is there a possibility to write a formula 
to calculate back the red and green channel intensities after normalisation 
of the data. Do I need to incorporate this formula in my R script?  I am 
biologist ans would seek your help to write this formula.

Thanking you,

Binita



Dr. Binita Dutta
MicroArray Facility(MAF)
UZ Gasthuisberg
Onderwijs en Navorsing
Herestraat 49
3000 Leuven
Belgium
Henrik Bengtsson
# 
Hi, here's the calculations

  M = log2(R/G)     = [log rules] = log2(R)-log2(G)     (1)
  A = 1/2*log2(R*G) = [log rules] = (log2(R)+log2(G))/2 (2)

where log2(x) is the logarithm of x with base 2. If R,G > 0, there is
a one-to-one relationship between (A,M) and (G,R) as follows

  R = (2^(2A+M))^(1/2) = sqrt(2^(2A+M))   (3)
  G = (2^(2A-M))^(1/2) = sqrt(2^(2A-M))   (4)

It's a good exercise to verify the correctness, by replacing R and G
in (1) and (2) with the expressions for R and G in (3) and (4).

I believe there are methods in the marray packages for calculating R
and G from M and A like the above, but I don't know the exact name of
them.

BTW: I think it's better to ask these type of questions to the
Bioconductor (of which marrayNorm is part of) mailing list instead.

Best wishes

Henrik Bengtsson
Lund University, Sweden
channel