Interaction snp * environment strata

Question

[mel22]

Hello, I would like to calculate same table as below, but I can't understand how they did it ? they don't show any reference strata ?! I am working on R Thank you for your help !

Answer 1

[Kevin Blighe]

Hi Melania - welcome.

It can be assumed that the reference level / stratum is Controls, and that the interaction variable is 'exposure', encoded as Nonexposed|Exposed. [to diagnostic radiation].

You can encode this in R using glm(). The y outcome variable should be Controls|Cases, with Controls being the reference level, and the independent predictors are the genotypes.

For example:

model <- glm(y ~ rs17301766 * exposure, data = mydata, family = binomial(link = 'logit'))
summary(model)

Kevin

[mel22]

Hi Kevin Thank you very much, As shown in the table they have OR for both Genotype subtypes , so which of these Genotypes was taken as reference as we see OR for both types homozygous major allelle and hete+homo minor allel ?

[Kevin Blighe]

Hi, the information is actually in the methods:

An unconditional logistic regression model was used to estimate the ORs and 95% CIs for association between exposure to diagnostic X-rays and risk of thyroid cancer and its subtypes in different genotypic strata. To increase statistical power, the heterozygous and homozygous variant genotypes were combined for each SNP and compared with the "common homozygous" genotype. The significance of gene–exposure interaction was assessed by adding an interaction term in the logistic models. Results were adjusted for age (continuous), gender, race, and body mass index (BMI).

[source: https://cebp.aacrjournals.org/content/cebp/early/2018/02/20/1055-9965.EPI-17-0319.full.pdf]

Note the final line in bold, too.

[mel22]

Thanks Kevin, I am sorry, I think my question was not clear. In fact, for each variable we have to consider a reference level. Do for genotypes they have two different level, on of theme is refernce. My question is how they have an OR for both genotype levels ?

Thanks again

[Kevin Blighe]

They probably just run each model twice, and flip the reference level for the genotype. That is what I assume.

[mel22]

Ah Ok ! I see, Many thanks for your help !