[BioC] tilingArray- normalizeByReference

Joern Toedling toedling at ebi.ac.uk
Tue Oct 28 11:07:41 CET 2008


Hello Anjan,

well, the error message indicates that your vector pm_vector is not
usable by the function. Is it really an integer vector. What is the
output of
str(pm_vector)
?
Probably, it would be better to read in the file by something like
pm_vector <- read.delim("PerfectMatchIndex", skip=<put number of header
lines here!>)$V1
Again, check the result using
str(pm_vector)
If you want to use other functionalities of package tilingArray, you
should consider creating a 'probeAnno' environment, which holds the
mapping between probes and genomic match positions, for your array.
Please see the script "makeProbeAnno.R" in the package's "scripts"
directory for details. Once you have such a probeAnno object, the
function "PMindex" can be used to obtain a numeric vector that you can
use as "pm" in the normalization function.
Please see the vignette "assessNorm" of package tilingArray for more
details.

Best regards,
Joern

Anjan Purkayastha wrote:
> hi all,
> a rookie tilingArray question.
> i'm having trouble creating the perfect match index vector.
> for my probe set (n= 6490), each probe is a unique, perfect match to
> its target.
> i have created a probe index file that contains the integers: 1,
> 2....6490 (one index per line).
> i scanned this file as:  pm_vector <- scan("PerfectMatchIndex", 0)
>
> i used the pm_vector in a following command:
> vac_normalized_exp_data <- normalizeByReference(vac_exp_set,
> vac_hyb_set, pm= pm_vector, background= bg_vector, cutoffQuantile= 0)
>
> this gives me an error:
> 'pm' must be an integer vector with values between 1 and 6490.
>
> any idea what the problem could be?
> tia
> anjan
>

-- 
Joern Toedling
EMBL - European Bioinformatics Institute
Wellcome Trust Genome Campus
Hinxton, Cambridge CB10 1SD
United Kingdom
Phone  +44(0)1223 492566
Email  toedling at ebi.ac.uk



More information about the Bioconductor mailing list