Tim, r-free flags are used to calculate maximum-likelihood parameters (sigmaa, or alpha/beta - depending on parameterization), as well as m and D in 2mFo-DFc and mFo-DFc maps. For this they are supposed to be evenly distributed across resolution range, and the number of flags per thin enough resolution slice (such that alpha/beta can be considered constants in such bin) should be sufficient for their calculation. Sufficient amount of flags per bin typically varies between 50-200 depending on specific implementation (50 in CNS I guess, 150 in phenix.refine). This requirement is the major one that dictates the choice of free-r flags, both their amount and distribution. If this does not hold refinement may not go well.
Pavel On Tue, Mar 26, 2013 at 1:33 AM, Tim Gruene <[email protected]> wrote: > Dear all, > > I recall that the set of Rfree reflections should be 500-1000, rather than > 5-10%, but I cannot find the reference for it (maybe Ian Tickle?). > > I would therefore like to be confirmed or corrected: > > Is there an absolute number required for Rfree to be significant, i.e. > 500-1000 > irrespective of the total number of unique reflections in the data set, or > is it > 5-10% (as a compromise)? > > Thanks and regards, > Tim > > -- > -- > Dr Tim Gruene > Institut fuer anorganische Chemie > Tammannstr. 4 > D-37077 Goettingen > > GPG Key ID = A46BEE1A > > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.10 (GNU/Linux) > > iD8DBQFRUV09UxlJ7aRr7hoRAvs1AKCH4dEZMOMvZNYLICzcpNCCgXKP0wCgk7I9 > 2Bd/O0oR+dBtDhZQqBgsTHY= > =U2kE > -----END PGP SIGNATURE----- > >
