I agree this 3:1, 4:1, ... 10:1 etc is a good approach - especially for screening - as it samples more concentration conditions than the plain vanilla 1:1 set-up. One generally good thing too is that the precipitant also starts at lower concentration - so less chance of setting up a load of trials and watching them immediately crash out (although obviously not a problem in this particular case!) .

I think the reason for differences from pre-prepared concentrated protein are often to do with other components that come in with the protein - so anything in the protein solution like buffer, salt, reductant, detergent, metal ions, etc will be concentrated up alongside the protein.

cheers
Martyn

Martyn Symmons, MBU Cambridge.

As an aside I heard that the 1:1 derives from the days when pipettes could not be trusted with such small volumes. The 1:1 ratio is the best chance to get a droplet with reproducible relative concentrations regardless of any systematic error in the exact volumes dispensed.

From: Jacob Keller <[log in to unmask]>
To: [log in to unmask]
Sent: Friday, 5 March, 2010 17:02:15
Subject: Re: [ccp4bb] crystallization of a macromolecular complex

20 g/L is the same as 20 mg/ml, isn't it? That does not seem particularly high to me.

Why not try 200 g/L?

As a variant to this, you could just try doing 10:1 prot:ppt drops, and allowing the protein concentration to increase in situ. In my limited experience, however, this is not the same as setting up drops 1:1 with more concentrated protein, for reasons unknown to me. I wonder whether that is the experience of others on the BB as well?

JPK