Correct.

I feel we're going a little off topic here but if you really must know, the data is medical and I already have a controlled ad-hoc environment with everything you suggest.

The purpose of this application is scientific. We're mining data and looking for statistical meaning where none was before. If a user works hard at creating a filter that will produce a list of brain scans with such and such characteristics belonging to people that show this and that symptom and are smokers or have a genotype like this or like that, you can probably guess that there are many reasons to save that query. I can think of two big ones: 1) perhaps the user will want to come back later, change a single parameter out of the 30 he already set and get a new filtered dataset from there without having to start over; 2) if that researcher publishes a science paper based on the data, I would imagine he'd want a little bit of traceability, i.e. be able to tell exactly where the data is coming from.

I've said enough about the what and the why, any idea how?