[BioC] marray layout

Jean Yee Hwa Yang jeany at maths.usyd.edu.au
Wed Jul 4 21:49:01 CEST 2007


Hi Vanessa,

maSub is used to "indicate" the spots that are in the maGb, maGf etc 
of your marrayRaw.  In your case, the full array has 21632 spots but 
probably only 21504 "real" spots with values.  Therefore, the maSub should 
be a logical vector of length 21632 indicating the location of the 21504.

i.e.
The maSub vector will have 21632 entries with 21504 TRUE and 128 FALSE. 
Correctly setting up this vector is important in getting the right spatial 
plot.  Usually, in the Gene Info file from the data base, there should be 
spot position information that helps you determine which spots from the 
"full grid" (or full array) that is missing.  Typical the last few spots 
from each sub-grid (print-tip group).

Have a go at setting it up and if you are still having trouble, let me 
know where the data file is and I will have a look at it.

Cheers

Jean


>
> The array has 21632 spots, however in the database file I only have
> information for 21504 spots.
> Moreover, some of the probes are not matchable to a gene and therefore
> for those I would like to
> neglect their expression values. I did this now by parsing the database
> file and setting the expression values
> at NA for those "non-matchable" probes.
>
> Hence I have 21504 probes, of which 801 are controls and 784 non-matchable
> (NAs in their expression values).
> I tried to declare this with maSub that I use only part of the array
> (21504 spots and not 21632).
> My maControls only considers the 21504 spots.
>
> I read in the whole thing and then look at the graphs created by one
> color "image". There seems to be a lot
> of white spots in my arrays and they are not consistent in the same
> place on the array ( although all the NAs
> refer to the same 784 probes in the 3 arrays, see code below).
> I was wondering if this has something to do with this maSub... and that
> it put these NAs (21632-21504, only about 100 ones) randomly on the array?
> *How can I correctly read in my marrayLayout?*
> I only want to use image to see if there is some spatial gradient in my
> arrays. If htere is, I need to use
> background substraction, otherwise I won't.
>
> This is the code I use...I am reading in 3 arrays of the format Chalfie379.
> In blue the output I get from R.
> The pdf file with the M images  can be found at
> http://www.psb.ugent.be/~vamei/Docs/Chalfie379_image.pdf
>
> library(limma)
> library(marray)
> library("Biobase")
>
> Chalfie379targets=read.marrayInfo("file_A-SMDB-379", info.id=c(2,4,5),
> labels=c(2))
> summary(Chalfie379targets)
> Object of class marrayInfo.
>
>   maLabels      SMDfile         green Cy3 CH1               red Cy5 CH2
> 1 new17040.xls new17040.xls wt sorted touch cell  mutant sorted touch cell
> 2 new17041.xls new17041.xls wt sorted touch cell  mutant sorted touch cell
> 3 new17042.xls new17042.xls wt sorted touch cell  mutant sorted touch cell
>
> Number of labels:  3 Dimensions of maInfo matrix:  3  rows by  3  columns
>
> Notes:
> file_A-SMDB-379
>
> file379 <- as.vector(Chalfie379targets at maInfo[,1])
> Chalfie379.gnames <- read.marrayInfo(file379[1],info.id=c(1,2,3), labels=2)
> summary(Chalfie379.gnames)
> Object of class marrayInfo.
>
> maLabels Spot    Name      ID
> 1     AC3.5    1   AC3.5   AC3.5
> 2     AC3.8    2   AC3.8   AC3.8
> 3     AH6.6    3   AH6.6   AH6.6
> 4   B0001.3    4 B0001.3 B0001.3
> 5   B0024.1    5 B0024.1 B0024.1
> 6   B0024.4    6 B0024.4 B0024.4
> 7   C06H5.1    7 C06H5.1 C06H5.1
> 8   C07A9.5    8 C07A9.5 C07A9.5
> 9   C07G2.1    9 C07G2.1 C07G2.1
> 10  C08B6.4   10 C08B6.4 C08B6.4
> ...
>
> Number of labels:  21504 Dimensions of maInfo matrix:  21504  rows by
> 3  columns
>
> Notes:
> new17040.xls
>
> Chalfie379.L <- new("marrayLayout", maNgr=8, maNgc=4, maNsc=26, maNsr=26)
> maControls(Chalfie379.L) <- Chalfie379.gnames at maInfo[,3]
> length(Chalfie379.L at maControls)
> S <- c(1:maNspots(Chalfie379.L))
> T <- Chalfie379.gnames at maInfo[,1]
> Z <- S %in% T
> maSub(Chalfie379.L) <- Z
> ctl <- rep("control", length(T))
> ctl[maControls(Chalfie379.L) != "control"] <- "probe"
> maControls(Chalfie379.L) <- factor(ctl)
> summary(Chalfie379.L)
> Array layout:    Object of class marrayLayout.
>
> Total number of spots:                  21632
> Dimensions of grid matrix:              8 rows by 4 cols
> Dimensions of spot matrices:            26 rows by 26 cols
>
> Currently working with a subset of 21504spots.
>
> Control spots:
> There are   2 types of controls :
>
> control   probe
> 801   20703
>
>
> Notes on layout:
>
> Chalfie379.raw <- read.marrayRaw(file379,name.Gf="CH1_I",
> name.Gb="CH1_B", name.Rf="CH2_I",
> name.Rb="CH2_B", gnames=Chalfie379.gnames, layout=Chalfie379.L,
> targets=Chalfie379targets, DEBUG=TRUE)
>
> pdf("Chalfie379_image.pdf")
> image(Chalfie379.raw, col=1, main="1")
> image(Chalfie379.raw[,2], col=1, main="2")
> image(Chalfie379.raw[,3], col=1, main="3")
> dev.off()
>
> myM <- Chalfie379.raw[,2]@maGf /Chalfie379.raw[,2]@maRf
> sum(is.na(myM))
> [1] 784
>
> > summary(which(is.na( Chalfie379.raw[,3]@maGf )))
> Min. 1st Qu.  Median    Mean 3rd Qu.    Max.
>  22    5564   10690   10770   15920   21450
> > summary(which(is.na( Chalfie379.raw[,2]@maGf )))
> Min. 1st Qu.  Median    Mean 3rd Qu.    Max.
>  22    5564   10690   10770   15920   21450
> > summary(which(is.na( Chalfie379.raw[,1]@maGf )))
> Min. 1st Qu.  Median    Mean 3rd Qu.    Max.
>  22    5564   10690   10770   15920   21450
>
> Thank you so much for your help!
>
> Best regards,
> Vanessa Vermeirssen
>
> -- 
>
> -- 
> ==================================================================
> Vanessa Vermeirssen, PhD
>
> Tel:+32 (0)9 331 38 23                        fax:+32 (0)9 3313809
> VIB Department of Plant Systems Biology, Ghent University
> Technologiepark 927, 9052 Gent, BELGIUM
> vamei at psb.ugent.be                         http://www.psb.ugent.be
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor
>



More information about the Bioconductor mailing list