[ccp4bb] Is there an easy way to down-weight figures of merit?

[William Scott]

Hi Citizens:

I realize this is one of those warranty-violating questions, but is  
there a simple way to down-weight figures of merit, or the equivalent  
in terms of Hendrickson-Lattman coefficients?

I'm trying to generalize and automate a procedure for solving RNA  
structures without heavy-atom phases by feeding phaser random-sequence  
model A-form RNA helices. It pretty much works, but there comes a  
point where I want to use the phase probability distribution from the  
partial/inaccurate model in solvent-flattening while discarding the  
model, with the hope of minimizing model bias.  The problem is that if  
the refinement of the partial/inaccurate model goes reasonably well,  
the FOM become inflated to the point where solvent-flattening becomes  
less effective for improving and correcting phases.

Thanks.

Bill Scott