Good afternoon,

I have a doubt regarding a Cox model with interaction items.

I have the following variables:

gender - male/female
percentage x -low/high
gene y - continuous (gene expression of gene y)

I built the model with:

coxph(formula=Surv(PFI.t,PFI.e)=gender + percentage:gene_y, data=mydata,model=TRUE)

I then get the following summary of the model:

                               coef        exp(coef)  se(coef)    z   Pr(>|z|)  
genderMALE                     0.2958729   1.3442993 0.2234927 1.324   0.1855  
percentagehigh:gene_y          0.0064450   1.0064658 0.0026070 2.472   0.0134 *
percentagelow:gene_y           0.0007183   1.0007185 0.0057817 0.124   0.9011  
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

I have two questions:

1) Is my interpretation correct that with increasing gene expression of gene y the risk of dieing increases for the group high percentage

2) I would be interested to turn it around: I would like to see the results for decreasing gene expression of gene y in group high percentage and risk of dieing. How can I do that?

Thank you!