[BioC] HTqPCR

[Heidi Dvinge]

Hello Arun,

> Dear Heidi
> Iam using HTqPCR for my 384 well TLDA data analysis. Iam analysis 40
> patient sample and 20 normal samples. For my analysis is it possible
> to get the fold change of that 40 patient samples individually
> campared to mean of all normal. at present analysis iam getting mean
> fold change of all patient compared to all normal.
> if i can get that i can find which all patients have higher value than
> normal.
>
There's no default way of doing this in HTqPCR, but you can easily do it
'manually' by extracting the Ct values. If you haven't already done it,
you should consider removing unreliable Ct values first, to avoid them
biasing the results. For example:

# Load example data
data(qPCRpros)

# Mark unreliable values, if this isn't already done
q.tmp <- setCategory(qPCRpros, Ct.max = 35, Ct.min = 10, quantile=NULL)
q.filter <- filterCategory(q.tmp)
# Count number of NAs, just FYI
colSums(is.na(getCt(q.filter)))

# In this example, I assume that the first 3 samples are the controls
control.index <- sampleNames(q.filter) %in% c("sample1", "sample2",
"sample3")
# Get the mean of these
tmp <- getCt(q.filter)[,control.index]
control.mean <- rowMeans(tmp, na.rm=TRUE)

# Comapre to all the non-controls
tmp	<- getCt(q.filter)[,!control.index]
sample.vs.control <- tmp-control.mean

> secondly is it possible to get the excell table data of normalized ct
> values.
>
Yep, you can export these using the standard write.table() function:

# Get Ct values
tmp <- getCt(qPCRpros)
# Write to file
write.table(tmp, file="normalized_Ct_values.xls", sep="\t")

HTH
\Heidi
> Deepak
> + 91 471 2522338
>