Hi the reason could be that refmac converts all aniso B to iso B when you do overall B value refinement. Without data it would be difficult to test and give definite answer.
regards Garib On 9 Oct 2010, at 00:15, Hailiang Zhang wrote: > Hi, > > I found there are many changes between refmac_5.5 and refmac_5.2. For > example, the key word "REFI BREF OVER" will result in totally different > results under these 2 versions. Based on my input PDB with anisotropic B > pre-refined, refmac_5.5 gave a much higher R/Rfree than refmac_5.2. Can > somebody explain the difference and give me some suggestions on how to > modfiy the keyword in refmac_5.5 to match what refmac_5.2 have done by > using "REFI BREF OVER"? > > Thanks!
