Wilcoxon signed rank test help

Hi - I have data that I wish to perform this statistical test upon but there are a few things i am unsure about, i hope somebody can help!

The data we have is from about 30 patients - cancer tissue and normal tissue was sampled from each patient and we have determined the percentage methylation in a gene. So we have paired data (cancer vs normal) and a continous variable (not normally distributed) which we expect to be higher in the cancer samples. Perfect candidate for the wilcoxon signed rank test I believe?

The thing is that we have measured the methylation percentage of each tumour and normal sample three times, and there is some variability in the technique so we do not get the exact same measurement every time. So should I use the average of these three measurements as my input data for the statistical test? I am unsure whether this would be correct, and also how would the variability of the measurement technique be taken into account?

I am performing it in Rcmdr.

Thanks for any help you can provide!