[BioC] Re: Single channel normalization

[cmprobst]

I really appreciated you help, Natalie.=0D
=0D
I wasn't aware of the Single Channel cDNA functions and documentation that =
were included in the great package Limma.=0D
=0D
One final question remains.=0D
=0D
As I am working with just one channel, I have used normalizeBetweenArrays d=
irectly in a matrix of log transformed Cy3 intensities, with the quantile m=
ethod.=0D
I haven't found any description if normalizeBetweenArrays assumes log-trans=
form (or non logged) data, or if there is an option to turn it on or off.=
=0D
=0D
I presume there is no problem in using normalizeBetweenArrays directly on l=
og-transformed data, isn't it?=0D
=0D
TIA=0D
=0D
Christian=0D
=0D
De:Natalie Thorne =0D
=0D
Para:cmprobst =0D
=0D
C=F3pia:bioconductor at stat.math.ethz.ch, p.little at unsw.edu.au, djn at maths.uns=
w.edu.au, Gordon K Smyth , Natalie Thorne , TERRY Speed =0D
=0D
Data:Wed, 10 Sep 2003 14:29:22 +1000=0D
=0D
Assunto:Re: Single channel normalization=0D
=0D
  =0D
=0D
> =0D
> Hi Christian,=0D
> =0D
> We have just added to limma some functions and documentation for doing=0D
> single-channel normalization. Limma is an R package that has been include=
d=0D
> in the Bioconductor release package. Gordon Smyth is the maintainer=0D
> for limma and the package is frequently updated. You should go to=0D
> http://bioinf.wehi.edu.au/limma/ for the latest release which now include=
s=0D
> the single-channel(sc) normalization functionality and function for=0D
> plotting the single-channel densities.=0D
> =0D
> This latest release has also just been posted to Bioconductor to the=0D
> Developmental Package of limma at http://www.bioconductor.org/ and will b=
e=0D
> available there shortly.=0D
> =0D
> In R use help.start() and go to packages...limma...Accommpanying=0D
> documentation...usersguide (in .html or .pdf)=0D
> =0D
> Section 14 of the usersguide is on single-channel normalization for=0D
> two-color arrays. This gives an example of normalizing the=0D
> single-channels of the ApoAI data set which is part of sma. =0D
> See also the help examples for the function plotDensities,=0D
> normalizeWithinArrays, and normalizeBetweenArrays. Some of these=0D
> examples are given below.=0D
> =0D
> =0D
> library(limma)=0D
> library(sma)=0D
> help.start()=0D
> data(MouseArray)=0D
> # just bkg corr=0D
> MA.n <- MA.RG(mouse.data) # no normalization=0D
> =0D
> MA.q <- normalizeBetweenArrays(MA.p, method =3D "q") # quantile norm=0D
> =0D
> G.q <- normalizeBetweenArrays(RG.MA(MA.n)$G,method=3D"q") # only green sc=
's=0D
> # takes a matrix =0D
> =0D
> tmp<-cbind(RG.MA(MA.n)$R,RG.MA(MA.n)$G)[,c(1,3,8,9,12)] # select sc's=0D
> tmp.q <- normalizeBetweenArrays(tmp, method =3D "q") # takes a matrix =0D
> =0D
> MA.p <- normalizeWithinArrays(MA.n, mouse.setup) # default p-loess=0D
> MA.pq <- normalizeBetweenArrays(MA.p, method =3D "q") # pq norm=0D
> =0D
> MA.MpAq <- normalizeBetweenArrays(MA.p, method =3D "Aq") # MpAq norm=0D
> # Yang & Thorne 03=0D
> =0D
> =0D
> plotDensities(MA.n) # default=0D
> =0D
> plotDensities(MA.n,arrays=3Dc(1:6), # same as default=0D
> groups=3Dc(rep(1,6),rep(2,6)),col=3Dc("red","green"))=0D
> =0D
> plotDensities(MA.n,arrays=3DNULL,groups=3DNULL, # diff cols=0D
> col=3Dc("blue","purple"))=0D
> =0D
> plotDensities(MA.n,singlechannels=3Dc(1,2,7)) # indexing sc's=0D
> =0D
> plotDensities(MA.n,singlechannels=3Dc(1,2,7), # diff cols=0D
> col=3Dc("pink","purple"))=0D
> =0D
> plotDensities(MA.n,singlechannels=3Dc(1,2,7), # groups=3D2=0D
> col=3Dc("pink","purple","blue")) # col too long=0D
> =0D
> plotDensities(MA.n,singlechannels=3Dc(1,2,7), # controlling=0D
> groups=3Dc(1,2,3),col=3Dc("pink","purple","blue")) # col and groups=0D
> =0D
> plotDensities(MA.n,singlechannels=3Dc(1,2,7),groups=3Dc(1,1,1),=0D
> col=3Dc("purple"))=0D
> =0D
> # All single-channels, three groups (ctl,tmt,reference), three colors.=0D
> plotDensities(MA.n,singlechannels=3Dc(1:12),=0D
> groups=3Dc(rep(1,3),rep(2,3),rep(3,6)),col=3Dc("darkred","red","green"))=
=0D
> =0D
> # Densities after single-channel MpAq (preferred) normalization.=0D
> plotDensities(MA.MpAq)=0D
> =0D
> =0D
> * * Natalie Thorne, Phd Student.=0D
> * * =0D
> * * Genetics and Bioinformatics Division,=0D
> * The Walter And Eliza Hall Institute=0D
> * * Of Medical Research (WEHI). =0D
> * * PO Royal Melbourne Hospital, 3050=0D
> * * PH: +61 3 9345 2631=0D
> * Fax: +61 3 9347 0852 =0D
> * * URL http://bioinf.wehi.edu.au=0D
> * * =0D
> * * Mathematics and Statistics Department, =0D
> * The University of Melbourne =0D
> * * Fax: +61 3 9344 4599=0D
> * * =0D
> * * =0D
> * CRC for discovery of genes for common =0D
> * * human diseases.=0D
> * *=0D
> * * Email : thorne at wehi.edu.au=0D
> =0D
> On Tue, 9 Sep 2003, cmprobst wrote:=0D
> =0D
> > Hi, Natalie=0D
> > =0D
> > I would like to ask you some question about single channel normalizatio=
n, mainly the how-to.=0D
> > =0D
> > I am asking you for suggestion in how to plot the density functions as=
=0D
> was done in the paper with Jean Yang. =0D
> > =0D
> > And also, if it is possible to apply the quantiles.normalization (and i=
f=0D
> yes, how) in a matrix of m spots per n hybs. =0D
> > =0D
> > TIA=0D
> > =0D
> > Christian=0D
> =0D
> =0D
> =0D
> =0D
> =0D
> =0D
> Esta mensagem foi verificada pelo E-mail Protegido Terra.=0D
> Scan engine: VirusScan / Atualizado em 03/09/2003 / Vers=E3o: 1.3.13=0D
> Proteja o seu e-mail Terra: http://www.emailprotegido.terra.com.br/=0D
>=20
	[[alternative HTML version deleted]]