R factors and B factors

[Joe Krahn]

Here's my opinions on R & B factors.

alex jack wrote:
> 
> Dear All:
> 
> Could somebody help me answer these questions:
> 
> 1. What's generally the acceptable R-factors for a structure, with the
> data of 2.5 A; 3.0 A; and 3.5 A.
Generally low 20s is OK, less than 20 is good.  Sigma-cutoffs can
lower the working & free R a percent or two, at the expense of
producing a less accurate structure.
 
> 2. Is it true, that the higher resolution your data has, the lower
> R-factors you should get? Or just the way around?
Generally, they're about the same, mainly because low resolution
structures have a lot of small errors.  However, a low resolution
structure based on a high-resolution model (i.e. looking at a different
ligand state) should have lower R-factors.

> 
> 3. What's the significance of the atomic B-factors when you have a low
> resolution data, for example, 3.0 A; or 3.5 A.
>
Often, low resolution structures have high overall B-factors.  These
are due to variations in the molecular orientation throughout the
crystal, not due to thermal motion of atoms.  This is actually a major
contributor to B-factors in most macromolecular structures, which is
why many people dislike the term temperature-factor.