[BioC] diagnostics

[alex lam (RI)]

Dear Dr. Smyth,

I figured out what went wrong with my previous imageplot. In my data
files and gal file the rows were sorted by gene names. I re-sorted the
rows by spot number and reloaded the files. Now the images are now much
more even, and the streak noticed on the 5th array on the maanova
spatial plot is now visible.

http://www-gene.cimr.cam.ac.uk/todd/arrayview/new-image-M-1-6.png

Thanks,
Alex

------------------------------------
Alex Lam
PhD student
Department of Genetics and Genomics
Roslin Institute (Edinburgh)
Roslin
Midlothian EH25 9PS

Phone +44 131 5274471
Web   http://www.roslin.ac.uk


-----Original Message-----
From: alex lam (RI) 
Sent: 06 February 2006 17:44
To: 'bioconductor at stat.math.ethz.ch'
Cc: 'smyth at wehi.edu.au'; 'gary.churchill at jax.org'
Subject: diagnostics


Dear colleagues,

I loaded data from some cDNA arrays into limma and maanova/R separately
and was checking if my arrays were ok. I ran imageplot in limma and
arrayview in maanova, and to my surprise the images looked very
different.

I have put the images of the first six arrays on
http://www-gene.cimr.cam.ac.uk/todd/arrayview/

image-M-1-6.png is the limma output, the others png files came from
maanova.

Are they not plotting the same thing (log ratio)? I didn't get any
background values so I plotted M. Arrayview005.png shows a streak which
is not visible in QPG185 (same array) in limma. And I don't understand
why there are regular blocks of almost solid red or green in the limma
output. I didn't remove landlites and controls (but gave them 0 weight
in limma) but they are supposed to be regularly spaced. And the images
from maanova differ a lot in how red or green they are. Should I worry
about the quality of these arrays? 

Limma code:
MA<-normalizeWithinArrays(RG)
imageplot3by2(MA, z="M")

Maanova code:
data.raw<-read.madata(...)
arrayview(data.raw, array=1:6)

Any help would be greatly appreciated.
Many thanks,
Alex

------------------------------------
Alex Lam
PhD student
Department of Genetics and Genomics
Roslin Institute (Edinburgh)
Roslin
Midlothian EH25 9PS

Phone +44 131 5274471
Web   http://www.roslin.ac.uk