Title: Message
Hi List
Members
First of all, my
apologies if this gets posted twice. I originally submitted the email 4
hours ago and I haven't yet seen it appear on the list.
On to my
question:
A common method of
analysis at my company is to take a binary result from one experiment and
compare it with the result from another - I'll diagram below. Each "gene"
is independent with respect to the other genes in the experiment. I
have heard that these tables can be analyzed for a null hypothesis using
Fisher's Exact Test - can Stata be configured to do this?
Example
Outcomes:
Gene ! Exp1 !
Exp2
-----+------+------
2001 !
1 ! 1
2002 !
1 ! 1
2003 !
1 ! 0
2004 !
1 ! 0
2005 ! 1
! 1
2006 !
1 ! 1
2007 ! 0 ! 1
2008 ! 0 ! 0
2009 !
1 ! 1
2010 ! 0
! 0
Concordance
Table
Exp1
1
0 Sum
---+----+----+----
!
! !
E 1! 5
! 1 ! 6
x !
! !
p--+----+----+----
2 ! !
!
0! 2
! 2 ! 4
! ! !
---+----+----+----
S !
! !
u ! 7 ! 3 ! 10
m
! ! !
As I understand
it, the null hypothesis is the chance that the table could appear in its
observed configuration by chance through random sampling.
Thanks for any
guidance
John Wallace
Research
Associate
Affymetrix, Inc
(408)731-5574
3380 Central Expressway 609C
Santa Clara, CA 95051
---
Outgoing mail is certified Virus Free.
Checked by AVG
anti-virus system (http://www.grisoft.com).
Version: 6.0.449 / Virus
Database: 251 - Release Date: 1/27/2003
