I want to run some epigenome wide association studies based on city and pollutant type, could you please suggest if I can get DNA methylation sites (CpG) association based on pollutant type for each city. like below I am trying to find Pb toxic elements association site in four different cities which has varying level of Pb. we do have 10 toxic element so not sure should I run model with one toxic element at a time or should include all in this model?

```
#model matrix
var<-model.matrix(~log2Pb*City + sex + Age +CD8T +CD4T +NK + Bcell +Mono, data=data1)
fit<-lmFit(mval,var)
fit2<-eBayes(fit,trend=TRUE, robust=TRUE)
probe<-topTable(fit2,adjust="BH",coef=???,num=Inf)
```

try to run interaction model using limma but not sure if this is the right approach and what to mention in coef????

Thanks for your all help!

Jim, there are 4 cities so the interaction will be 3 coefficients.

Indeed. Thanks for pointing that out!