On Thu, 26 Jan 2006 08:39:36 -0800 (PST), "Nel N. (sent by
Nabble.com)" <lists at nabble.com> wrote:
>>hi-i'm a modelling(math) student and are working on cancer predictions which uses biology dataset-so i dun quite hav biology knowledge-however,the dateset i'm working on which was obtained from Affymetrix chip(micrarray technology), it contained genes expressions values which include positive and negative values-i've tried searching the net for the meaning of those values but fail to get an exact answer-to what i know, the values should represent some intensity of presence of genes-if so, how come there's negative values?i mean aren't intensity values supposed to be all positive??
>i hope someone can enlighten me with their microarray knowledge
You'll need to check with those who provided the data, but it is
likely that the results are based on comparison with a control. (eg.,
cancer vs normal). In that case, differences can be pos or neg. As
someone already noted, a log scale might be involved.
Note that the test is not for _presence_ of a gene, but for its
_expression_. So a test condition (e.g, cancer) may have higher or
lower expression. (Actually, gene presence may be altered in cancer
cells, but I suspect that is a side issue here.)
I hope you are working closely with biologists on this. That should
add tremendously to the understanding that might come from such