[phenixbb] Phaser question : RMS deviation and FOM

[Randy J. Read]

Dear Peter,

You're absolutely right that it would be better if Phaser refined the 
Sigma-A values at the end of the molecular replacement job. This is on our 
to-do list, and will be done at some point. It shouldn't be all that hard, 
but so far something else always seems to be more urgent. In the meantime, 
you can usually take the solution into refinement and, even if you just do 
rigid-body refinement, phenix.refine will compute FOMs and maps with 
refined error estimates. However, phenix.refine can't deal with ensembles, 
which is where there's still a role for Phaser.

Thanks for the comment!

Randy Read

On Nov 6 2009, Peter Grey wrote:

>Dear Phenix developers,
>
>The expected RMS deviation input to Phaser affects the FOM estimation of
>phases quality.
>This in-turn affects the sigmaa weighted maps and most importantly the
>density modification steps that follow phaser.
>I would like to ask if, based on the phaser solution and the LLG parameter,
>one could optimize the RMS estimation or know the real RMS deviation that
>would result in more true FOM.
>
>Thanks,
>
>Peter
>
>With condolences to the family of Warren DeLano who will be sorely missed
>