Question: Logistic regression with PLINK, p-values
0
gravatar for monsegd14
6.3 years ago by
monsegd140
Spain
monsegd140 wrote:

We want to evaluate the association of several SNPs in an population with a disease. After the bech work we have a database with the results. We use the software PLINK to run logistic regression analysis including covariates like the sex of the patientcs. The output file give me several p-values :

 

My questions is: Which p-value give me the association between the SNP and the disease? which p-value reflect the effect of the covariate?

snp • 5.3k views
ADD COMMENTlink modified 6.2 years ago • written 6.3 years ago by monsegd140

You'll need to either upload an image of that table somewhere or simply make a table in your post and manually type in the values.

ADD REPLYlink written 6.3 years ago by Devon Ryan97k

The output file is:

SNP Allele Test OR SE L95 U95 P
1 c ADD 0.2 0.3 0.1 0.3 3.728e-05
1 t ADD 4.9 0.3 2.8 8.8 3.728e-05
1 C SEX 1.5 0.3

0.7

3.2 0.325
1 T SEX 1.8 0.6 0.4 7.1 0.382
1 C ADDxSEX 1.1 0.4 0.4 2.5 0.825
1 T ADDxSEX 0.9 0.4 0.3 2.2 0.825

 

ADD REPLYlink written 6.3 years ago by monsegd140

Thank you so much for your help. I was totally lost with this kind of analysis because I thought that the ADD p-value was for SNP-disease interaction (the p-value is significant) and SEX p-value for Sex-disease association (not significant) but I do not understad the ADDxSEX colum (not significant and higher tan SEX-p-value) I mean with a P-value so high does it mean that SEX has influenced in the disease?? (because when SEX effect is added, the p-value change from significant to not significant)

I am totally lost

 

ADD REPLYlink written 6.2 years ago by monsegd140
1

Gender has an insignificant p-value, so no, it doesn't seem to influence anything here. "ADDxSEX" is looking at a possible interaction between the disease and gender, for which there also appears to be none. You might want to have a local statistician walk you through what a logistic regression is and what an interaction means.

ADD REPLYlink written 6.2 years ago by Devon Ryan97k

Thank you son much again...but the relation between the SNPs-disease corrected by sex, which analyse or p-value is??

Sorry for my ignorance, and thank you for your time and your help

ADD REPLYlink written 6.2 years ago by monsegd140
1

ADD is the SNP disease result while accounting for gender.

ADD REPLYlink written 6.2 years ago by Devon Ryan97k
0
gravatar for chrchang523
6.3 years ago by
chrchang5237.4k
United States
chrchang5237.4k wrote:

In your output file, ADD p-values are for SNP-disease association, SEX p-values are for sex-disease association, and ADDxSEX is for the multiplicative interaction term.

ADD COMMENTlink written 6.3 years ago by chrchang5237.4k

Thank you!!!!!

I want to apply the same analyse with the age of the patients, but I have problems to upload age in plink. I have tried to code the age: 0 if the age is <60 and 1 if the age is >60

 

Thank you!!!!

ADD REPLYlink written 6.2 years ago by monsegd140
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 2.3.0
Traffic: 985 users visited in the last hour