Hi all:
I'm new to the group and hoping to get some advice on proc genmod which I a=
m=
 using to analyze correlated data. Rather than multiple measurements per =
patient over time, I have tests for multiple genotypes conducted at the sam=
e=
 time.

I am trying to model the probability of genotype1 being positive in Native =
=
Americans (vs all other races) while controlling for the presence of other =
=
related genotypes. Except for msmt (and id, of course), all variables are =
=
dichotomous (0,1). Each measurement is actually a test result for genotypes=
 =
2 through 4. My data look like this:

ID         NA        msmt    geno1   geno2   geno3   geno4
1          1          1          1          1          0          0
1          1          2          1          0          0          0
1          1          3          1          0          0          0
2          0          1          0          1          0          0
2          0          2          0          0          1          0
2          0          3          0          0          0          1
....
9706     0          1          0          0          0          0
9706     0          2          0          0          1          0
9706     0          3          0          0          0          1

And my code:

proc genmod data=3Dall descending order=3Ddata;
class NA ID ;
model geno1=3DNA geno2 geno3 geno4/dist=3Dbinomial
link =3D logit ;
repeated subject=3DID /type=3Dexch;
run;

Have I set up my data and code properly=3F

Thanks in advance for assistance and apologies in advance if this is the =
wrong forum....



 Theresa Mickiewicz, MSPH
Research Analyst
Denver Health: Public Health Informatics Group
605 Bannock St., Mail Code 2600
Denver, CO 80204
Office: 303-602-3686 (new)
=46ax: 303-602-3676 (new)


The Denver Health email system has made the following annotations

---------------------------------------------------------------------

CONFIDENTIALITY NOTICE - This e-mail transmission, and any documents, files=
 =
or previous e-mail messages attached to it may contain information that is =
=
confidential or legally privileged.  If you are not the intended recipient,=
 =
or a person responsible for delivering it to the intended recipient, you ar=
e=
 hereby notified that you must not read this transmission and that any =
disclosure, copying, printing, distribution or use of any of the informatio=
n=
 contained in or attached to this transmission is STRICTLY PROHIBITED.  If =
=
you have received this transmission in error, please immediately notify the=
 =
sender by telephone or return e-mail and delete the original transmission =
=
and its attachments without reading or saving in any manner.  Thank you.

---------------------------------------------------------------------

Theresa,

For the three IDs for which you presented example data,
the variable NA has the same value as geno1.  Why is that?
What does NA measure?

If NA is redundant with geno1 throughout your data, then
your effort to understand the relationship between geno1
and the other three genotypes will be futile.  NA would
explain geno1 without information from any of the other
genotypes.

One other question comes to mind.  Why are you treating
geno1 as the response?  Is it truly the genotype of
interest or might you turn around and use geno2 as the
response variable with predictors geno1, geno3, and
geno4?  If geno1 is not THE response variable of
interest, then I would approach things differently.

Dale

---------------------------------------
Dale McLerran
Fred Hutchinson Cancer Research Center
mailto: dmclerra@NO_SPAMfhcrc.org
Ph:  (206) 667-2926
Fax: (206) 667-5977
---------------------------------------


--- On Wed, 11/18/09, Mickiewicz, Theresa <Theresa.Mickiewicz@DHHA.ORG> wrote:

> From: Mickiewicz, Theresa <Theresa.Mickiewicz@DHHA.ORG>
> Subject: correlated data/proc genmod
> To: SAS-L@LISTSERV.UGA.EDU
> Date: Wednesday, November 18, 2009, 1:10 PM
> Hi all:
> I'm new to the group and hoping to get some advice on proc
> genmod which I am using to analyze correlated data. Rather
> than multiple measurements per patient over time, I have
> tests for multiple genotypes conducted at the same time.
>
> I am trying to model the probability of genotype1 being
> positive in Native Americans (vs all other races) while
> controlling for the presence of other related genotypes.
> Except for msmt (and id, of course), all variables are
> dichotomous (0,1). Each measurement is actually a test
> result for genotypes 2 through 4. My data look like this:
>
> ID     NA      msmt   geno1   geno2   geno3   geno4
> 1      1        1       1       1       0       0
> 1      1        2       1       0       0       0
> 1      1        3       1       0       0       0
> 2      0        1       0       1       0       0
> 2      0        2       0       0       1       0
> 2      0        3       0       0       0       1
> ...
> 9706   0        1       0       0       0       0
> 9706   0        2       0       0       1       0
> 9706   0        3       0       0       0       1
>
> And my code:
>
> proc genmod data=all descending order=data;
> class NA ID ;
> model geno1=NA geno2 geno3 geno4/dist=binomial
> link = logit ;
> repeated subject=ID /type=exch;
> run;
>
> Have I set up my data and code properly?
>
> Thanks in advance for assistance and apologies in advance
> if this is the wrong forum....
>
>
>
>  Theresa Mickiewicz, MSPH
> Research Analyst
> Denver Health: Public Health Informatics Group
> 605 Bannock St., Mail Code 2600
> Denver, CO 80204
> Office: 303-602-3686 (new)
> Fax: 303-602-3676 (new)
>
>
> The Denver Health email system has made the following
> annotations
>
> ---------------------------------------------------------------------
>
> CONFIDENTIALITY NOTICE - This e-mail transmission, and any
> documents, files or previous e-mail messages attached to it
> may contain information that is confidential or legally
> privileged.  If you are not the intended recipient, or
> a person responsible for delivering it to the intended
> recipient, you are hereby notified that you must not read
> this transmission and that any disclosure, copying,
> printing, distribution or use of any of the information
> contained in or attached to this transmission is STRICTLY
> PROHIBITED.  If you have received this transmission in
> error, please immediately notify the sender by telephone or
> return e-mail and delete the original transmission and its
> attachments without reading or saving in any manner.
> Thank you.
>
> ---------------------------------------------------------------------
>