Juergen Bosch [[log in to unmask]] wrote
> As a rule of thumb you should expect R values around your
> resolution/10 so in the low to mid 20 - 18/23 sound reasonable
> for that resolution.
I'm as guilty as anyone for promulgating that Rule of Thumb,
but I have since realized the error of my ways.
Please find attached a plot of the R and Rfree distribution of all
PDB entries as a function of resolution between 1 and 4 A,
with the erroneous Rule of Thumb superimposed.
As you will see, it holds only in a very restricted region around
2.0A resolution. Even there the slope is not a good match.
Ethan "gnuplot maniac" Merritt
|