[BioC] RLE Nose t2

noemi noemi.andor at campus.lmu.de
Fri Jun 18 00:22:19 CEST 2010

James W. MacDonald <jmacdon at ...> writes:

> Hi Noemi,
> Noemi Andor wrote:
> > Hi,
> > 
> > I have used the RMA-Express upon my .CEL files and after that the
> > NOSE RLE T2 test to filter out bad quality arrays. Yet I have so many
> > cel-files that I can't allocate them to their t2-value (they can't be
> > distiguished in the picture). Instead I can print the RLE and NOSE
> > statistics separatly with their corresponding.CEL file. Now my
> > question is, how can the t2 statistic be calculated from the NOSE and
> > RLE statistic? Or are there any other solutions to allocate bad
> > arrays?
> Please take a look at the posting guide for hints how to ask questions:
> http://www.bioconductor.org/docs/postingGuide.html
> Maybe others on this list know what a NOSE RLE T2 test is, but it would 
> be helpful for those of us who don't if you were to state the package 
> you are using, possibly the code (and any errors) and a sessionInfo().
> Otherwise, you are expecting people to do a bunch of work to decipher 
> your question, and very few have the time or inclination to do so.
> Best,
> Jim
> > 
> > thank's,
> > 
> > Noemi
> > 
> > _______________________________________________ Bioconductor mailing
> > list Bioconductor at ... 
> > https://stat.ethz.ch/mailman/listinfo/bioconductor Search the
> > archives:
> > http://news.gmane.org/gmane.science.biology.informatics.conductor

ok, well I used the HT_HG_U133A cdf -file and had about 386 .cel-files to
normalize whit RMA, for which I used RMA-express. After normalizations some
statistics were available to help one decide which arrays/ cel-files had bad
the RLE statistic: RLE = Run Length Encoding (I think)
and the NOSE statistic
The T2 RLE NOSE statistic combines the two statistics into one single number,
which is plottet for every cel-file. Yet there are to many cel-files to read
them from the x-achsis of the plot, so I have to figure out some other way to
associate each cel file with it's corresponding statistic. As there is an option
to print the NUSE and RLE statistics separatly, I can use them to calculate the
so called "t2 nuse RLE" statistic:
So what I have is:
cel-file:	1	2....
min	a1
LQ	a2
mean	a3
UQ	a4
max	a5

cel-file:	1	2....
min	b1
LQ	b2
mean	b3
UQ	b4
max	b5

What I want is:
cel-file	1	2...
t2		?

part of the log-file:
Reading in data
Opening CDF and CEL files
Done Reading in datafiles

Choose Preprocessing Steps
Carrying out Analysis
Background Adjusting.
Normalizing Using Quantile Normalization.
Summarizing Using PLM.
Done computing expression measure using PLM.

Visualizing QC statistics

I couldn't find that information anywhere, so I thought some of you who used
RME-express, have some hints.



More information about the Bioconductor mailing list