[phenixbb] phenix and weak data

Ed Pozharski epozh001 at umaryland.edu
Mon Dec 10 14:25:22 PST 2012


On Mon, 2012-12-10 at 13:17 -0800, Pavel Afonine wrote:
> The point was that one always needs to draw a line between
> pedanticism 
> and practicality. Otherwise we would be all going back from FFT to 
> direct summation in Fcalc and gradients calculation, as that would 
> surely improve R-factors by 0.5-1% or so in some cases through
> improving 
> the accuracy of refined model parameters.
Let's not change the subject to some impractical examples.  Optimal
resolution cutoff is a very practical question that has nothing to do
with FFT/direct summation.

In practice, resolution cutoff is always been selected based on a
particular set of prejudices individual crystallographer subscribes to.
Karplus&Diederichs came up with a very reasonable approach for objective
selection of such cutoff.  It is simple and very practical and leads to
(admittedly minor) model improvement.  What's not to like?

My objection is to you moving the bar to "significant changes in map
quality" level.  It's unreasonable (there are very few things that do
that and your criterion itself is too vague).  My deeply seated mistrust
of the R-values notwithstanding, reasonable changes that can be proven
to result in better correlation between experimental data and model
predictions are to be welcomed.  


"Hurry up before we all come back to our senses!"
                           Julian, King of Lemurs

More information about the phenixbb mailing list