[Arrays] standard error of fold change when using a Universal Reference

Andrew McArthur via arrays%40net.bio.net (by amcarthur from mac.com)
Mon Jun 18 09:05:30 EST 2007


Hello all,

We have performed an experiment that included replicated microarray  
analyses.  We have 4 replicate RNA samples from three treatments (A,  
B, C) plus a Universal Reference RNA sample (UR) made up of equal  
amounts of all A, B, and C samples.

All experimental RNA samples (A1, A2, A3, A4, B1, B2, B3, B4, C1, C2,  
C3, C4) were Cy3 labeled and individually co-hybridized with Cy5  
labeled Universal Reference.  When all was said and done we  
calculated the mean and standard error for fold change of A, B, and C  
versus the Universal Reference.

Now we want to calculate "through" the Universal Reference to compare  
A versus B.  To calculate the mean fold change of A versus B we have  
simply divided the mean fold change of A against the UR by the mean  
fold change of B against the UR.  The question now is how to  
calculate the standard error of the fold change of A versus B.

Does anyone have a good reference for calculating the standard error  
of a ratio variable when the ratio is made of two means with their  
own standard errors?

Much thanks,
Andrew McArthur

------
Andrew G. McArthur, Ph.D.
Email: amcarthur from mac.com, Web: http://mcarthurlab.blogspot.com



-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.bio.net/bionet/mm/arrays/attachments/20070618/60d687a5/attachment.html


More information about the Arrays mailing list