[BioC] HTqPCR plotting groups of interest

[Simon Melov]

This worked great Heidi, 
thanks very much! It would be great if you could put this in the next update as an example. When you have a lot of groups, you dont want to plot them on every graph. Also, I think you have the color function set to  only plot 10 colors as a default. People might be confused by that when you have more than 10 groups. You end up with 10 distinct colors for the first 10 columns, and the rest of the bars plotted have no color, I just assigned the number of colors to the number of groups used)

thanks again!

best

s

On Jul 3, 2012, at 1:53 PM, Heidi Dvinge wrote:

> Sorry all, pressed "send" too soon.
> 
>> Hi Simon,
>> 
>>> If anyone can help, it would be great.
>>> 
>>> I want to plot different groups of interest in HTqPCR using some of the
>>> plot functions, and I've played around with plotting different ways a
>>> number of times, and cant get the plot functions in HTqPCR to display
>>> what
>>> I want. I've a large number of groups, and dont want to plot everything
>>> every time (plotting everything works just fine after Heidi's help).
>>> 
>>> #the plate6 CSV file links the sample ID to the group (Treatment) it
>>> belongs to#
>>>> plate6=read.csv("PlateG6.csv", header=TRUE, as.is=TRUE)
>>>> head(plate6)
>>>  Sample Treatment
>>> 1    NTC       NTC
>>> 2    L29       LFY
>>> 3    M10       MMO
>>> 4    C30       CFY
>>> 5    S15       SFO
>>> 6    S28       SMY
>>>> group6=as.vector(t(plate6$Treatment))
>>>> sample6=as.vector(t(plate6$Sample))
>>>> raw6=readCtData(files="Chip6NTC.csv", format="BioMark", n.features=48,
>>>> n.data=48, samples=sample6)
>>>> g=featureNames(raw6)[1]
>>>> g
>>> [1] "TUB"
>>>> plotCtOverview(raw6, genes=g, col=rainbow(48), group=group6,
>>>> Legend=FALSE)
>>> #works fine, plots the first gene Tubulin nicely, however, now I want to
>>> compare my NTC group with my LFY group (listed in "Treatment" above)#
>> 
>> Do you just want the Ct values of NTC relative to LFY? In that case have a
>> look at the 'calibrator' parameter. As shown in the plotCtOverview
>> examples, setting calibrator="LFY" will show all Ct values for other
>> groups with LFY subtracted first.
>> 
>>> #I've tried the below to just plot one group, the "NTC" group, and it
>>> doesnt work (I'm not clear why, whatever is being plotted is not
>>> correct),
>>> #
>>> #not sure how to plot just two groups with this function (or three, or
>>> however many I want)#
>>> 
>>>> plotCtOverview(raw6, genes=g, col=rainbow(48), group=group6[1],
>>> legend=FALSE)
>>> 
>> The problem here is that you're inputting an object with *multiple*
> columns, but
>> you're only providing link from 1 column to a group. On top of that,
>> plotCtOverview has no way of knowing which of the six columns you mean.
>> 
>> For plotting just some of your groups, you therefore need to subset your
>> qPCRset object. For example:
>> 
> index <- group6  %in% c("SFY", "LFY", "CFY")
> plotCtOverview(raw6[,index], genes=g, col=rainbow(48),
> group=group6[index], Legend=FALSE)
> 
> NOw you're limiting your data to these 3 groups, but you still have a
> column-to-group mapping for all your data.
> 
> HTH
> \Heidi
> 
>> 
>> 
>>> thanks in advance
>>> 
>>> Simon
>> 
>> 
> 
>