how to generate random size matched background regions

[gene_x]

Guys,
Many papers do this kind of enrichment of a particular factor (TF or histone peaks) at their regions of interest. They typically get a randomly drawn, size matched control regions to estimate the background distribution. I only know how to draw a group of fixed-length random background regions using bedtools.

However, the regions of interest are typically of variable length. In this case, how do you get a random group of background regions with the same size distribution?

Thanks!

[biocomputer]

This is an old question but I'll reply here instead of making a new post. There's another old unanswered question asking basically the same thing, and also mentions choosing regions with respect to sequence bias (GC, repets, etc.) http://seqanswers.com/forums/showthread.php?t=3011.

I have thought that it might be a good idea to use your ChIP-seq input sample to guide selection of the random regions. The reads in your input will be biased "correctly" for GC, repetitive sequences, unmapped regions, etc., similar to the IP sample. So if you need say, 1000 random regions, what about choosing 1000 random aligned reads from your input sample and using those as the start coordinates for each region? The end coordinates would depend whether you want all regions the same size or variable.

Does this sound like a reasonable strategy? If not, what's the best way to choose random regions to compare to my ChIP-seq peaks?