non-parametric tests with significant difference in sample sizes I have outcomes from a biological expt., where we have measured DNA lengths that has been affected. 
For condition1, I have 11 observations (length of DNA) and for condition2 I have approx. 116000. I want to compare for difference in median lengths between these two conditions. I have tried wilcox.test in R, but I would like to know if what would be the proper way to check for differences since the sample size is way too huge for condition2. I think simple wilcox.test would not be sufficient in this case. Is there any method which would take the sample sizes into account before testing ? 
Would it wise to do bootstrapping for condition2, and compare with condition1 for 1000 (n) times to check how the p-values differ each case ?
Thank you. 
 A: 
I have outcomes from a biological expt., where we have measured DNA lengths that has been affected.
For condition1, I have 11 observations (length of DNA) and for condition2 I have approx. 116000. 
I want to compare for difference in median lengths between these two conditions. 

Without additional assumptions, the Wilcoxon-Mann-Whitney is not a test for equality of medians.

I have tried wilcox.test in R, but I would like to know if what would be the proper way to check for differences since the sample size is way too huge for condition2.

Not so. There's really no problem with using this test with one large sample.

Would it wise to do bootstrapping for condition2, and compare with condition1 for 1000 (n) times to check how the p-values differ each case ?

You don't need anything so complex. You could do a straight bootstrap test but why not simply do a permutation test (or with such a large second sample, a randomization test)?

Reagrding the data, it is highly skewed, and hence I believe it should be analysed non-parametrically. 

"Highly skew" doesn't imply "analyze non-parametrically". It might suggest not making a parametric assumption of normality, though. Not that I would advise against a nonparametric test, just that your reasoning is telling you the data is not normal, not that all parametric tests will be unsuitable.
