[BioC] qPCRnorm

[Heidi Dvinge]

Hello Andreia,


> Dear Heidi,
>
> thanks for the message, why stratify="type" doesn't work with
> plotCtOverview??.

plotCtOverview was only designed to be used for a handful of genes
individually across multiple samples. Using e.g. all Target or Endogenous
Controls would often either result in hundreds of genes being plotted.
Alternatively, values could be summarised by  taking the average across 
genes within each sample, but that probably wouldn't be so informative.

Instead, plotCtOverview() comes with the parameter "genes" for selecting
the genes of interest. This can either be i) a list of gene names present
in featurenames, ii) a vector of the genes to choose, such as
c(1,6,60,100), or iii) a TRUE/FALSE vector of the same length as number of
features. So if you want t select e.g. all Endogenous Controls, a possible
appraoch would be to say:

> data(qPCRraw)
> g	<- featureType(qPCRraw)=="Endogenous Control"
> plotCtOverview(qPCRraw, genes=g, xlim=c(0,10))


> I also sent you the email because I am getting error
> when
> I try to see the help files:
> ?filterCtData
> Error in print.help_files_with_topic(x) :
>   No text help for 'filterCtData' is available:
> corresponding file is missing
> In addition: Warning message:
> In
> print.help_files_with_topic("C:/PROGRA~1/R/R-29~1.2/library/HTqPCR/chm/filterCtData")
> :
>   No CHM help for 'filterCtData' in package 'HTqPCR' is available:
> the CHM file for the package is missing
>
Hm, this definitely looks like some sort of windows-related issue. Can you
perhaps send the exact commands you type in starting from loading the
package, along with the output of sessionInfo()? Is this just the case for
this one function, or does it apply to all functions in HTqPCR?

Cheers
\Heidi

> Thanks
> Andreia
>
> On Mon, Jan 11, 2010 at 6:07 PM, Heidi Dvinge <heidi at ebi.ac.uk> wrote:
>
>> Hello Andreia,
>>
>> sure, if you want to remove some of the data points in HTqPCR you can
>> use
>> the function filterCtData(). If featureType of your genes are e.g.
>> "Endogenous Control" and "Target", you can remove all the target genes;
>> filterCtData(qPCRset, remove.type=c("Target")). See ?filterCtData for
>> more
>> examples.
>>
>> Note that for many of the plotting functions this is actually not
>> necessary, since you can automatically stratify the data based on
>> various
>> feature characteristics. If you're interested in different featureType()
>> or
>> featureClass(), you can for example say plotCtBoxes(qPCRset,
>> stratify="type") or with stratify="class". For other plot types, such as
>> scatter plots, features can be coloured differently depending on
>> featureType
>> or some other characteristics; e.g. plotCtScatter(qPCRset, col="type").
>>
>> Cheers
>> \Heidi
>>
>>
>>
>> On 11 Jan 2010, at 16:37, Andreia Fonseca wrote:
>>
>>  Hello Heidi,
>>>
>>> thanks for the code, until now everything seems to be o.k. I am trying
>>> to
>>> filter data and I was wondering if you have created functions to select
>>> data, the reason is that I want to plot the Ct values just for the
>>> endogenous control targets to see if there are significant differences
>>> between treatments, something like select.type??
>>> Cheers
>>> Andreia
>>>
>>> On Fri, Jan 8, 2010 at 4:45 PM, Heidi Dvinge <heidi at ebi.ac.uk> wrote:
>>> Hello Andreia,
>>>
>>> I've attached a file with all the latest source code. I still need to
>>> add
>>> some things plus delete other and correct the documentation, before I
>>> compile it together into a package. However, in your case it should
>>> work
>>> if you say:
>>>
>>> > library(HTqPCR)   # To get the vignette and help files available
>>> > source("where.ever.you.place.the.file/source_functions_v1.1.1.R") #
>>> Overwrite old functions with corrected ones.
>>>
>>> I don't have access to a windows machine right now so I can't test it,
>>> but
>>> if you still get errors when you try to run through the vignette, then
>>> drop me a line. The sooner I get bugs corrected, the better :)
>>>
>>> Cheers
>>> \Heidi
>>>
>>> > Hi Heidi,
>>> >
>>> > can you send me the sample data file, this way I can read it  and
>>> > continue.
>>> > cheers
>>> > Andreia
>>> >
>>> > On Fri, Jan 8, 2010 at 11:37 AM, Heidi Dvinge <heidi at ebi.ac.uk>
>>> wrote:
>>> >
>>> >> Hi Andreia,
>>> >>
>>> >> well, sadly enough that's because I'm a numpty!! When I wrote the
>>> >> function(s), rather than defining file position using file.path, I
>>> just
>>> >> concatenated names with "/" because at that point I was only
>>> intending
>>> >> on
>>> >> using this function for myself on my mac.
>>> >>
>>> >> On mac/linux this can be fixed for readCtData by having the data you
>>> >> want
>>> >> to read in a different directory than your current working
>>> directory.
>>> >> However, for windows you'll need a bug-fixed version of
>>> readCtData(). I
>>> >> have this available, but it hasn't been submitted to the BioC devel
>>> >> repository yet - it's creeping rapidly higher up on my to-do list.
>>> If
>>> >> you
>>> >> want, I can send you a version off-list.
>>> >>
>>> >> Cheers
>>> >> \Heidi
>>> >>
>>> >> > Hi Heidi,
>>> >> >
>>> >> > Thanks for the advise. I am going through the HTqPCR documentation
>>> but
>>> >> I
>>> >> > am
>>> >> > getting the following error while trying to read the sample input
>>> >> data:
>>> >> >
>>> >> > path <- system.file("exData", package = "HTqPCR")
>>> >> >> head(read.delim(file.path(path, "files.txt")))
>>> >> > Error in file(file, "r") : cannot open the connection
>>> >> > In addition: Warning message:
>>> >> > In file(file, "r") :
>>> >> >   cannot open file '/files.txt': No such file or directory
>>> >> >> files <- read.delim(file.path(path, "files.txt"))
>>> >> > Error in file(file, "r") : cannot open the connection
>>> >> > In addition: Warning message:
>>> >> > In file(file, "r") :
>>> >> >   cannot open file '/files.txt': No such file or directory
>>> >> >
>>> >> >
>>> >> > I am working in a windows machine and using R 2.9.2
>>> >> >
>>> >> > thanks
>>> >> > Andreia
>>> >> >
>>> >> > On Fri, Jan 8, 2010 at 8:44 AM, Heidi Dvinge <heidi at ebi.ac.uk>
>>> wrote:
>>> >> >
>>> >> >> Hello Andreia,
>>> >> >>
>>> >> >> I'm not familiar with qpcrBatch objects, but <shameless self
>>> plug>
>>> if
>>> >> >> you're interested in testing for differential expression between
>>> >> genes,
>>> >> >> you could also consider the HTqPCR package. The raw data is read
>>> into
>>> >> a
>>> >> >> qPCRSet object, which is similar to ExpressionSets used for
>>> >> microarray
>>> >> >> data. The package performs various normalisations of the raw qPCR
>>> >> data,
>>> >> >> similar to qpcrNorm, along with some data visualisation and
>>> >> filtering.
>>> >> >> Differential expression can be analysed using a standard t-test
>>> or
>>> >> with
>>> >> >> a
>>> >> >> limma-based approach, or if you want to do something else than
>>> that
>>> >> you
>>> >> >> can extract all the values from your qPCRSet object using
>>> exprs().
>>> >> >> </shameless self plug>
>>> >> >>
>>> >> >> Cheers
>>> >> >> \Heidi
>>> >> >>
>>> >> >> > Dear list,
>>> >> >> >
>>> >> >> > I am trying to understand how qpcrNorm works, so I followed the
>>> >> >> > documentation, so I understood how to normalize the data, but
>>> now
>>> I
>>> >> >> want
>>> >> >> > to
>>> >> >> > test which genes are differentially expressed between batches
>>> and
>>> >> make
>>> >> >> a
>>> >> >> > Mann-Whitney U-test. How can I transform the normalized data
>>> which
>>> >> is
>>> >> >> a
>>> >> >> > object class qpcrBatch into a data.frame. Or else how can test
>>> >> using
>>> >> >> this
>>> >> >> > kind of object.
>>> >> >> > Thanks for the help
>>> >> >> > Andreia
>>> >> >> >
>>> >> >> >       [[alternative HTML version deleted]]
>>> >> >> >
>>> >> >> > _______________________________________________
>>> >> >> > Bioconductor mailing list
>>> >> >> > Bioconductor at stat.math.ethz.ch
>>> >> >> > https://stat.ethz.ch/mailman/listinfo/bioconductor
>>> >> >> > Search the archives:
>>> >> >> > http://news.gmane.org/gmane.science.biology.informatics.conductor
>>> >> >> >
>>> >> >>
>>> >> >>
>>> >> >>
>>> >> >
>>> >>
>>> >>
>>> >>
>>> >
>>>
>>>
>>
>
>
> --
> --------------------------------------------
> Andreia J. Amaral
> Unidade de Imunologia Clínica
> Instituto de Medicina Molecular
> Universidade de Lisboa
> email: andreiaamaral at fm.ul.pt
>          andreia.fonseca at gmail.com
>