Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.
|
|
|
|
Re: How to support with statistical parameter?
Posted:
Aug 17, 2012 10:03 PM
|
|
On Fri, 17 Aug 2012 08:02:33 -0700 (PDT), arulthasan83cyr@gmail.com
wrote:
>I have a list of over expressed genes from 15 experiments for same disease. I prepared a gene list, which are present in these all 15 groups, 14 groups, 13 groups, 12 groups, etc.
>I concluded that the genes which are present in all 15 groups are very important and i think this conclusion is appropriate in biological point of view. If i want to prove statistically that the genes which are present in 15 experiments are important.What type of Test, i have to apply?
The main thing that I know about the statistics used in
genetics studies is that those statisticians do not intersect
very much with those of us who do clinical research.
There are special analyses used for family studies.
There are special analyses used for micro-arrays with
2000 or 4000 genes.
I don't know if any readers here know enough to help you.
If you do not get a response here, you might look for a
Usenet group that is more specialized, or look for a Mail-List.
In case it may help here or elsewhere, I have a couple of
questions, even though a geneticist might know what you
are talking about.
Is "over expressed genes" a reference to something rather
unique? Or is "over" just a typo for "overt"? ("Overt" does
sound more innocent to me.)
Does an "experiment" with a "group" use one person? Is there
any Control? - If you don't have information about how often
a gene is *not* expressed, you don't have much chance of
drawing any conclusions.
--
Rich Ulrich
|
|
|
|
