[Bioperl-l] problem with not transformed negative numbers in .soft file of GEO database

[Remo Sanges]

On 02-Jun-04, at PM 05:22, Gong Wuming wrote:

> Hi list,
>
> I am afraid the question is not suitable to this mailing list, but I 
> can not find better place to post it.
>
> I have a problem with the .soft file in GEO database. Taking the 
> GDS169.soft file for example, what does the negative entry mean? How 
> to deal with these negative numbers before log transformation?
>

Gong,

the best place in wich to post this questions is:
bioconductor www.bioconductor.org or the
gene-array list http://www.gene-chips.com/gene-arrays.html

In your case the data you are observing should come from the old
version of Affymetrix algorithm (MAS4) that produced negative
expression value when the signal of negative-control probes was
higher than the signal of positive probes...

You can solve setting all negative value to zero, but keep in mind
that if you need to make statistical analysis on these values this is 
not
the best solution.

Remo