This has been discussed before, I guess more than once.... I think most people (I'm sure i'll be corrected if wrong) would favor not removing any atoms or setting occupancies to zero and let the invisible atoms be accounted for by high B-factors (either set manually or just letting refinement do its thing).
Bert ________________________________ From: CCP4 bulletin board <CCP4BB@JISCMAIL.AC.UK> on behalf of Vipul Panchal <panchal.vi...@igib.in> Sent: 04 May 2017 16:12 To: CCP4BB@JISCMAIL.AC.UK Subject: [ccp4bb] Poor density fit. HI all. I am solving protein structure with 2.16A resolution. There are two chain in an asymmetric unit. I see that in one of the chain, many residues' density for side chains is incomplete and therefore results in poor density fit. I want to know your opinions for the approach I have taken. Figures relevant to each approach have been attached herewith. Case1: There is no experimental density at all. Therefore, i have deleted side chains to Gly. Case2: Though there is incomplete density for Leu, it is enough to suggest its rotamer. In this case, as may be seen, i have just set occupancy for atoms without density(CG, CD1, CD2) to zero. Hopeful for the response. -- Vipul Panchal Senior Research Fellow, Respiratory disease and biology, CSIR-IGIB (M)-9540113372
