Dear Jiamu, On Fri, Jul 31, 2009 at 12:00:25AM +0800, Jiamu Du wrote: > I am refining a structure of a complex between of 50kD protein and a 20kD > glycosylated protien. The data is of 2.9 A resolution. The wilson B factor > is as high as 86.3 A^2. > The refinement seems well with R/Rf of 0.21/0.25. But the B-factor is extra > high. For the 50kD part, the average B factor is 76.5 A^2. But the B factor > of the 20 kD glycosylated protein is as high as 133.3 A^2. Although the > electron density looks fine, even the sugar chain is seen clearly. > My question is: > 1. How to reduce the B factor to a reasonable level?
How do you define 'reasonable'? And why would you want to reduce this anyway? ( 50*76.5 + 20*133.3 ) / 70 = 92.7 which seems fairly close to the Wilson B of 86.3, right? > 2. If it can not be redueced, when I published it, is this value acceptable? Better question: is it the right value? Remember it is results -> publish and not publish <- results ;-) If they're correct than they are acceptable (I would accept those values). > 3. In the same of similar resolutionIs, is there some other structures like > this situation? A component or a subunit of the protein has a extra high B > factor as high as 130. I'm sure hundreds of them ... but I'm sure you want your structure to stand on its own, so don't look too close at other structures and repeat the various mistakes we've all made and that are now set in stone in some old PDB file. Cheers Clemens -- *************************************************************** * Clemens Vonrhein, Ph.D. vonrhein AT GlobalPhasing DOT com * * Global Phasing Ltd. * Sheraton House, Castle Park * Cambridge CB3 0AX, UK *-------------------------------------------------------------- * BUSTER Development Group (http://www.globalphasing.com) ***************************************************************
