Hi there,
I would like to explore the idea of creating a function that is
a modification of the normalizeWithinArrays() with printtiploess
option of
the limma package.
We are spot quality control prior to normalization which results in
some
genes being dropped. Once that happens then we no longer have complete
sub-grids; normlizeWithinArrays will nor longer work with the printtip
loess option. So I wish to make a modification to the
normalizeWithinArrays
function that will re-index each sub-grid, so that the print top loess
can
be performed.
So I have been looking for the source code for limma, but it seams to
be
unavailable since the package is distributed as a binary. I have
checked
the author's page but its not there either.
I would like permission to look into the limma package source code and
re-factor the normalizeWithinArrays() function to do what I want.
Can somebody send me the source code ?
Peter
The code is available to you right now. The easiest way to get it is
to
type normalizeWithinArrays at an R prompt, copy and paste into your
favorite editor. If you want the source code with any comments that
may
exist, you will have to download the source version of limma, untar
and
gunzip it and then open norm.R in your favorite editor.
HTH,
Jim
James W. MacDonald
Affymetrix and cDNA Microarray Core
University of Michigan Cancer Center
1500 E. Medical Center Drive
7410 CCGC
Ann Arbor MI 48109
734-647-5623
>>> Peter Wilkinson <pwilkinson@videotron.ca> 07/14/04 05:36PM >>>
Hi there,
I would like to explore the idea of creating a function that is
a modification of the normalizeWithinArrays() with printtiploess
option of
the limma package.
We are spot quality control prior to normalization which results in
some
genes being dropped. Once that happens then we no longer have complete
sub-grids; normlizeWithinArrays will nor longer work with the printtip
loess option. So I wish to make a modification to the
normalizeWithinArrays
function that will re-index each sub-grid, so that the print top loess
can
be performed.
So I have been looking for the source code for limma, but it seams to
be
unavailable since the package is distributed as a binary. I have
checked
the author's page but its not there either.
I would like permission to look into the limma package source code and
re-factor the normalizeWithinArrays() function to do what I want.
Can somebody send me the source code ?
Peter
_______________________________________________
Bioconductor mailing list
Bioconductor@stat.math.ethz.ch
https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
Ah yes you are right I can type the function in. I would prefer to get
the
source code for the comments. The problem is that I did not notice
where
the source code resides. I have downloaded from the source link from
the
bioconductor site, and I will try that.
Thanks for your input.
Peter
At 10:44 AM 7/15/2004, James MacDonald wrote:
>The code is available to you right now. The easiest way to get it is
to
>type normalizeWithinArrays at an R prompt, copy and paste into your
>favorite editor. If you want the source code with any comments that
may
>exist, you will have to download the source version of limma, untar
and
>gunzip it and then open norm.R in your favorite editor.
>
>HTH,
>
>Jim
>
>
>
>James W. MacDonald
>Affymetrix and cDNA Microarray Core
>University of Michigan Cancer Center
>1500 E. Medical Center Drive
>7410 CCGC
>Ann Arbor MI 48109
>734-647-5623
>
> >>> Peter Wilkinson <pwilkinson@videotron.ca> 07/14/04 05:36PM >>>
>Hi there,
>
>I would like to explore the idea of creating a function that is
>a modification of the normalizeWithinArrays() with printtiploess
>option of
>the limma package.
>
>We are spot quality control prior to normalization which results in
>some
>genes being dropped. Once that happens then we no longer have
complete
>
>sub-grids; normlizeWithinArrays will nor longer work with the
printtip
>
>loess option. So I wish to make a modification to the
>normalizeWithinArrays
>function that will re-index each sub-grid, so that the print top
loess
>can
>be performed.
>
>So I have been looking for the source code for limma, but it seams to
>be
>unavailable since the package is distributed as a binary. I have
>checked
>the author's page but its not there either.
>
>I would like permission to look into the limma package source code
and
>
>re-factor the normalizeWithinArrays() function to do what I want.
>
>Can somebody send me the source code ?
>
>Peter
>
>_______________________________________________
>Bioconductor mailing list
>Bioconductor@stat.math.ethz.ch
>https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
On Thu, Jul 15, 2004 at 03:51:53PM -0400, Peter Wilkinson wrote:
>
> Ah yes you are right I can type the function in. I would prefer to
get the
> source code for the comments. The problem is that I did not notice
where
> the source code resides. I have downloaded from the source link from
the
> bioconductor site, and I will try that.
>
> Thanks for your input.
R packages are basically distributed in two formats: source and
binary.
Unix versions are always source and windows are always binary. As long
as you are only interested in the R code, you can - as James said -
always get the source from within R (although it might be difficult
depending on what function you want access to. "getAnywhere()" is nice
to know about"). I think (99% sure) that you cannot get the C code
from
within R. The thing to do is retriving the source tarballs. Eg.
getting
the source code for biobase may be done by pointing your webbrowser
at "www.bioconductor.org" , click "devel packages", "biobase" and
finally "source package download". Standard R packages may be gotten
from CRAN in a similar fashion.
But if you are simply looking for the R code, the thing to do is
probably accesing it through R.
/Kasper
> At 10:44 AM 7/15/2004, James MacDonald wrote:
> >The code is available to you right now. The easiest way to get it
is to
> >type normalizeWithinArrays at an R prompt, copy and paste into your
> >favorite editor. If you want the source code with any comments that
may
> >exist, you will have to download the source version of limma, untar
and
> >gunzip it and then open norm.R in your favorite editor.
> >
> >HTH,
> >
> >Jim
> >
> >
> >
> >James W. MacDonald
> >Affymetrix and cDNA Microarray Core
> >University of Michigan Cancer Center
> >1500 E. Medical Center Drive
> >7410 CCGC
> >Ann Arbor MI 48109
> >734-647-5623
> >
> >>>> Peter Wilkinson <pwilkinson@videotron.ca> 07/14/04 05:36PM >>>
> >Hi there,
> >
> >I would like to explore the idea of creating a function that is
> >a modification of the normalizeWithinArrays() with printtiploess
> >option of
> >the limma package.
> >
> >We are spot quality control prior to normalization which results in
> >some
> >genes being dropped. Once that happens then we no longer have
complete
> >
> >sub-grids; normlizeWithinArrays will nor longer work with the
printtip
> >
> >loess option. So I wish to make a modification to the
> >normalizeWithinArrays
> >function that will re-index each sub-grid, so that the print top
loess
> >can
> >be performed.
> >
> >So I have been looking for the source code for limma, but it seams
to
> >be
> >unavailable since the package is distributed as a binary. I have
> >checked
> >the author's page but its not there either.
> >
> >I would like permission to look into the limma package source code
and
> >
> >re-factor the normalizeWithinArrays() function to do what I want.
> >
> >Can somebody send me the source code ?
> >
> >Peter
> >
> >_______________________________________________
> >Bioconductor mailing list
> >Bioconductor@stat.math.ethz.ch
> >https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor@stat.math.ethz.ch
> https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor
--
Kasper Daniel Hansen, Research Assistant
Department of Biostatistics, University of Copenhagen
Hi Peter,
The source code of limma is certainly available, at:
http://bioinf.wehi.edu.au/limma/limma_1.7.2.tar.gz
You can unzip and untar the package with:
tar zxvf limma_1.7.2.tar.gz
or
gunzip limma_1.7.2.tar.gz
tar xvf limma_1.7.2.tar
If you are using Windows, you can obtain tar and gunzip from
the Rtools bundle at:
http://www.murdoch-sutherland.com/Rtools/
(Or you may be able to unzip a tarball with Winzip, but I'd
recommend using tar and gunzip.)
If you are using Mac, tar and gunzip should be available in the
apple developer tools.
Hopefully you realize that normalizeWithinArrays can already
accept weights, so you can give some spots zero weight if you
want, then exclude them completely after normalization.
One comment about typing the name of an R function and pressing
enter is that sometimes (especially on Windows) if the source
code of the function contains an escpaed character like a tab:
sep="\t"
it may be displayed as:
sep=" "
when you type the function name and press enter within R.
For an example of this, try typing the name of the limma
function read.maimages and pressing Enter, then scroll back up
and look at the 'sep' argument.
Regards,
James
Others have pointed out to you how to access the source code, but
there is
no real need to make the modification to limma that you seeking to do.
Instead of removing problem genes from the data set, simply assign
them
weight zero. This is exactly equivalent to removing them in the
normalization step but avoids having to re-index the sub-grids.
Gordon
> Hi there,
>
> I would like to explore the idea of creating a function that is
> a modification of the normalizeWithinArrays() with printtiploess
option
> of
> the limma package.
>
> We are spot quality control prior to normalization which results in
some
> genes being dropped. Once that happens then we no longer have
complete
> sub-grids; normlizeWithinArrays will nor longer work with the
printtip
> loess option. So I wish to make a modification to the
> normalizeWithinArrays
> function that will re-index each sub-grid, so that the print top
loess can
> be performed.
>
> So I have been looking for the source code for limma, but it seams
to be
> unavailable since the package is distributed as a binary. I have
checked
> the author's page but its not there either.
>
> I would like permission to look into the limma package source code
and
> re-factor the normalizeWithinArrays() function to do what I want.
>
> Can somebody send me the source code ?
>
> Peter
Hi there,
Thank you for your input. In fact this morning I tried just that,
setting
the gene to 0, and I found that to work.
Peter
At 09:57 PM 7/15/2004, Gordon K Smyth wrote:
>Others have pointed out to you how to access the source code, but
there is
>no real need to make the modification to limma that you seeking to
do.
>Instead of removing problem genes from the data set, simply assign
them
>weight zero. This is exactly equivalent to removing them in the
>normalization step but avoids having to re-index the sub-grids.
>
>Gordon
>
> > Hi there,
> >
> > I would like to explore the idea of creating a function that is
> > a modification of the normalizeWithinArrays() with printtiploess
option
> > of
> > the limma package.
> >
> > We are spot quality control prior to normalization which results
in some
> > genes being dropped. Once that happens then we no longer have
complete
> > sub-grids; normlizeWithinArrays will nor longer work with the
printtip
> > loess option. So I wish to make a modification to the
> > normalizeWithinArrays
> > function that will re-index each sub-grid, so that the print top
loess can
> > be performed.
> >
> > So I have been looking for the source code for limma, but it seams
to be
> > unavailable since the package is distributed as a binary. I have
checked
> > the author's page but its not there either.
> >
> > I would like permission to look into the limma package source code
and
> > re-factor the normalizeWithinArrays() function to do what I want.
> >
> > Can somebody send me the source code ?
> >
> > Peter
Hi
I'm using limma for an affymetrix dataset. For summarize which is the
most relavant differential expressed genes I use classifyTestsF() and
I extract up-regulated and down-regulated genes. At the same time i
look at the results of the fit with topTable() (sorted by B). But the
genes are not the same. Is it possible? Why?
What is the better solution, to summarize my results, between B and
F-statistics?
Thanks
Daniela Marconi
