[BioC] presence/absence call using oligo array?

[Benjamin Otto]

Hi all,

just as a matter of interest and for own understanding: On which base have
the oligos on the chip been designed? Can you take it for sure that these
oligos hybridize, in genetic theory, better with the parasite RNA than with
certain host RNA? I didn't ever work with long oligo arrays, so I can't
estimate how big this danger is in this case. So building upon the three
cases Sean mentioned and the ranking by fold change, maybe it would
iluminate the case a little bit more to identify the oligo sequence of genes
which seem interesting and blast it against the host genome, or at least
what is known of it. Maybe this would give a little better impression of how
big the effect of cross hybridization for this specific gene oligos has
been. Does this sound sensible?

regards,
Benjamin


> -----Original Message-----
> From: bioconductor-bounces at stat.math.ethz.ch
> [mailto:bioconductor-bounces at stat.math.ethz.ch]On Behalf Of Sean Davis
> Sent: 13 January 2006 19:03
> To: xpeng; a.mikhail at abdn.ac.uk
> Cc: Bioconductor
> Subject: Re: [BioC] presence/absence call using oligo array?
>
>
>
>
>
> On 1/13/06 12:49 PM, "xpeng" <xpeng at utk.edu> wrote:
>
> > Hi Amy,
> >
> > Thanks for the reply.
> >
> > You are right. The uninfected tissue is a blank, in theory. But the real
> > problem is that some probes cross-hybridize with house RNAs. I
> do not know the
> > oligo sequence yet. They know that there are more host RNAs
> than parasite RNAs
> > in the sample from the infected tissue, but not sure the percentage.
> >
> > Yes, they are 2-color arrays. The problem here is what to
> compare. It was
> > suggested to find genes with intensities higher than their own
> backgrounds,
> > say 4-fold higer, but I doubt it. I appreciate more inputs on
> the consequences
> > if it is done this way.
>
> Signals above background in the infected sample could be due to at least
> three possibilities:
>
> 1)  Hybridization by the parasite RNA
> 2)  Cross-hyb by the host RNA in the uninfected state
> 3)  Additional cross-hyb by host RNA due to the infection (upregulation of
> genes not highly expressed in the uninfected state).
>
> I agree with Amy that the best you can probably do is to look for
> differentially-expressed genes between your uninfected host and
> the infected
> sample.  I also agree with Amy that using absolute intensities as
> a measure
> of "expression" is probably not possible with most existing long-oligo
> technologies (although some new technologies like illumina seem to be
> working somewhat for that).
>
> Sean
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
>