[BioC] questions about Affy package from new user

[James MacDonald]

1.) Once you have done the summary, you don't have probe set data
anymore, so you cannot normalize on the probe set level. However, if you
are doing the mas5 algorithm the normalization (such as it is) occurs
after summarization.

2.) The normalization *is* done at the probe set level, so I don't
understand the question.

3.) What's GS? Well, regardless, you probably would not want to
re-normalize the data.

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

>>> "Lizhe Xu" <lxu at chnola-research.org> 03/14/04 06:00PM >>>
I started to use Bioconductor recently and had several questions about
the Affy package. Please help me and even answer to one question will be
appreciated. I know some question may take la ong paragraph to answer. 
(1) is it possible to do the summary first followed by normalization on
probe set level with Affy?
(2) what is the advantage to do normailization first, then probe set
summary compared to normalization at probe set level?
(3) After running bgcorrect, normalization and summary on probe set in
Affy (expresso function), I want to export the probe set data and
analyze it with GS (is there another package can do the same job as GS
in bioconductor)? Should I do the per chip normalization again in GS?
Thanks.
 
Lizhe

	-----Original Message----- 
	From: bioconductor-request at stat.math.ethz.ch
[mailto:bioconductor-request at stat.math.ethz.ch] 
	Sent: Sun 3/14/2004 5:04 AM 
	To: bioconductor at stat.math.ethz.ch 
	Cc: 
	Subject: Bioconductor Digest, Vol 13, Issue 26
	
	

	Send Bioconductor mailing list submissions to
	        bioconductor at stat.math.ethz.ch 
	
	To subscribe or unsubscribe via the World Wide Web, visit
	       
https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor 
	or, via email, send a message with subject or body 'help' to
	        bioconductor-request at stat.math.ethz.ch 
	
	You can reach the person managing the list at
	        bioconductor-owner at stat.math.ethz.ch 
	
	When replying, please edit your Subject line so it is more
specific
	than "Re: Contents of Bioconductor digest..."
	
	
	Today's Topics:
	
	   1. LC-MS data (Nicholas Lewin-Koh)
	
	
	----------------------------------------------------------------------
	
	Message: 1
	Date: Sat, 13 Mar 2004 22:48:50 +0800
	From: "Nicholas Lewin-Koh" <nikko at hailmail.net>
	Subject: [BioC] LC-MS data
	To: bioconductor at stat.math.ethz.ch, S.Nyangoma at cs.rug.nl 
	Message-ID:
<1079189330.2264.182616685 at webmail.messagingengine.com>
	Content-Type: text/plain; charset="ISO-8859-1"
	
	Hi,
	To my knowledge there are only 2 packages in R specifically for
MS data,
	mscalib on CRAN, and PROcess in bioconductor devel. The first is
for
	MALDI tof spectrometers and assumes you have picked peaks
already and
	works on the peaks list. The second is for seldi, but the
baseline
	correction and peak picking are pretty generic. To process LC-MS
data you
	have to decide how far back in the device internal processing
you want to
	go. Personally, I have found that the mantra for mass spec data
at the
	moment is "Don't trust vendor software". It mostly sucks. If you
can get
	it you want to be grabbing the data stream as it is read of the
column by
	the sensor, because it helps to warp the chromatagram from each
scan so
	that the peaks align properly. Then you want to conver to m/z.
After that
	comes all the signal processing song and dance, to subtract the
chemical
	noise, make a baseline adjustment, etc. The tools for this in R
are here
	and there and development for processing this stuff is nacent.
There is
	much more available in matlab, which though much more expensive
is mostly
	faster than R. The signal processing community and the
chemometrics
	people tend to work in matlab.
	
	
	Note that it has been my experience that automated peak
detection is an
	art, with more pitfalls than clustering. If you can do anything
to avoid
	that using prior knowledge it helps. Good luck.
	
	Nicholas
	>
	> Message: 2
	> Date: 12 Mar 2004 19:12:32 +0100
	> From: Stephen Nyangoma <S.Nyangoma at cs.rug.nl>
	> Subject: [BioC] LC-MS proteomics data
	> To: bioconductor at stat.math.ethz.ch 
	> Message-ID: <1079115152.10700.12.camel at iwi142>
	> Content-Type: text/plain
	>
	> Sorry for bothering you with this question.
	>
	> Has someone analylsed LC-MS data? How do you read this data
into R? Are
	> there preprocessing tools in R? What are the crusial
preprocessing
	> steps? Do the ascii files obtained from Brucker software
contain raw
	> files? Thanks. Stephen.
	>
	> 
	>
	>
	>
	
	
	
	------------------------------
	
	_______________________________________________
	Bioconductor mailing list
	Bioconductor at stat.math.ethz.ch 
	https://www.stat.math.ethz.ch/mailman/listinfo/bioconductor 
	
	
	End of Bioconductor Digest, Vol 13, Issue 26
	********************************************