The "rule" is that the average isotropic temperature factor for your protein atoms should not differ too wildly from the Wilson B-factor of the data set (when you convert your anisotropic B's to isotropic B's). Otherwise there is no such thing as an acceptable or unacceptable value, the value is inherent to the data set you refined with. You will find out that refinement programs have lower and upper limits (in the case of isotropic temperature factor refinement - in particular because there is a format constraint in pdb files, and also if the temperature factor is extremely high then there is no contribution of that atom to the diffraction signal, except at extremely low resolution). For anisotropic temperature factor refinement, I haven't checked the latter point, that the values are indeed limited to a certain range.
Fred. > Message du 07/07/10 18:12 > De : "Rakesh Joshi" > A : CCP4BB@JISCMAIL.AC.UK > Copie à : > Objet : [ccp4bb] HIGH B ANISOTROPIC VALUES > > > HI ALL, > > Thanks Dirk and Pavel for clearing up the discrepancy between the R values > going from Refmac5 to SF-check. On the same note, i have a large anisotropic > B factor( atomic model) once I carry out the TLSANL process. Is there an > acceptable value for this? > > cheers > rakesh > >
