[BioC] SCAN-UPC: 12079 of 22283 probes returned for hgu133a CELs?

[Steve Piccolo]

Hi Andrew,

Thanks for your email!

The normalization process in SCAN is done at the probe level. This is
executed in the same way, behind the scenes, regardless of whether a
probeSummaryPackage is specified. Thus at the probe level, the results
should be identical either way. However, the values that are returned by
the SCAN function are summarized at the probeset or gene level. When you
use the probeSummaryPackage option (such as when you're using BrainArray
annotations), it will do two things: 1) use a subset of probes as defined
in the annotations and 2) map the probes to gene IDs rather than
Affymetrix-defined probesets.

One of the great things about SCAN is that you can also get the underlying
(normalized) probe values and do whatever you want with them (e.g., apply
JetSet). To do this, specify a value for the probeLevelOutDirPath
parameter. This will create an output file in the specified directory that
contains the X-Y position and value for each probe.

Please let us know if any of your questions weren't answered.

Regards,
-Steve

---------------------------------------------------------------------------
----------------------------
Stephen Piccolo, Ph.D.
Postdoctoral Research Associate
@stevepiccolo


Affiliations:
  Department of Pharmacology and Toxicology, University of Utah
  Computational Biomedicine Section, Boston University School of Medicine
---------------------------------------------------------------------------
----------------------------




From:  Andrew Yates <yates.115 at osu.edu>
Date:  Tuesday, May 14, 2013 2:16 PM
To:  Stephen Piccolo <stephen.piccolo at hsc.utah.edu>
Cc:  "bioconductor at r-project.org" <bioconductor at r-project.org>
Subject:  Re: SCAN-UPC: 12079 of 22283 probes returned for hgu133a CELs?


Hi Bioconductor mailing list and Dr. Stephen Piccolo,

In SCAN.UPC, does the application of a custom probe summary package like
Brainarray[1] substantially change the normalized value per probe, or does
it merely return a subset of results? That is, what is the difference,
besides
 the number of probes returned, between "M.brainarray.SCAN" and
"M.all.SCAN" below?

Further, is it acceptable to apply SCAN.UPC with no probeSummaryPackage
parameter and then select representative probes using a package like
JetSet[2] afterwards?

Best,
Andrew

[1] 
http://brainarray.mbni.med.umich.edu/Brainarray/Database/CustomCDF/16.0.0/e
ntrezg.asp
[2] 
http://www.cbs.dtu.dk/biotools/jetset/
<http://www.cbs.dtu.dk/biotools/jetset/>

library(SCAN.UPC)
library(hgu133ahsentrezgprobe)
library(pd.hg.u133a)

M.brainarray.SCAN = SCAN(my.path.to.cels,
probeSummaryPackage=hgu133ahsentrezgprobe)

M.all.SCAN = SCAN(my.path.to.cels)





On Mon, Mar 4, 2013 at 4:50 PM, Andrew Yates
<yates.115 at osu.edu> wrote:

It does, thank you.


On Mon, Mar 4, 2013 at 1:39 PM, Steve Piccolo
<stephen.piccolo at hsc.utah.edu> wrote:

Hi Andrew,

Thanks for your email. The results you are seeing are expected.

The goal of the BrainArray mappings is to identify high-quality probes and
to enable mapping of individual probes to genes (rather than Affymetrix
probesets). On HG-U133A, BrainArray indicates that only 12,079 genes have
an adequate number of high-quality probes to generate a reliable
summarized value. Please let me know if that doesn't answer your question.

Regards,
-Stephen



From: Andrew Yates <yates.115 at osu.edu>
Date: Friday, March 1, 2013 10:15 PM
To: Stephen Piccolo <stephen.piccolo at hsc.utah.edu>
Subject: SCAN-UPC: 12079 of 22283 probes returned for hgu133a CELs?


Hi Dr. Piccolo,
I used SCAN-UPC to normalize arrays generated using the Affy hgu133a
platform using the probe annotations from the Brainarray Database[1]. Of
22,283 features in the platform, the resulting file from SCAN-UPC only had
 12,079. In downstream analysis, the results generated from these features
look good, but...

What accounts for these missing features, and did my selection of the
Brainarray Database have an effect on this? Did I use SCAN-UPC in the
recommended way? My R code is included below.

Best,
Andrew

http://brainarray.mbni.med.umich.edu/Brainarray/Database/CustomCDF/16.0.0/e
ntrezg.asp 
<http://brainarray.mbni.med.umich.edu/Brainarray/Database/CustomCDF/16.0.0/
entrezg.asp>

library(SCAN.UPC)
library(hgu133ahsentrezgprobe)
library(pd.hg.u133a)

LIU.SCAN = SCAN(fname.ptn, probeSummaryPackage=hgu133ahsentrezgprobe,
outFilePath="LIU.SCAN.txt")
LIU.UPC = UPC(fname.ptn, probeSummaryPackage=hgu133ahsentrezgprobe,
outFilePath="LIU.UPC.txt")