[BioC] Working with GERP scores

[Paul Leo]

Can anyone advise of an elegant solution for working with Genome wide GERP scores and R.

I would like to use them in coverage like calculations. Not sure how large they are as a run-length encoded object.


Thinking that:

Could prefilter to just regions I want. 
OR
Or restrict to GERP score thresholds  (But Filtered to bed like formats as text files they are @53Gb for Gerp score >1)
OR
Would an Rsql based solution work better?
OR 
Put them in vcf format .. compress then and use Rsamtools-like vcf tools readers.

Anyone tried any of these , what works?

Thanks 
Paul