Dear All, I am trying to solve a 2.6 A protein complex data and found the density for ligand but due to resolution constrains, it is not so sharp but clearly suggesting the fitting of ligand. The ligand is not so potent and tight binding inhibitor of enzyme and mimics the substrate and may exist in covalently linked acylated (single ring), hydrolyzed (single ring) and non-covalent (double ring) forms. Although deacylation rate of inhibitor is very slow in comparison to substrate therefore enzyme-inhibitor complex is expected to be in covalently linked state but the presence of other tow species in small proportion is also inevitable. The density is continuos through catalytic Ser residue to the carbonyl carbon atom of ligand which was expected to participate in covalent bond formation with enzyme. Although while fitting the ligand, it is always moving beyond the C-O covalent bond distance and forcing it by fixing atom resulted in distorted ligand geometry after refinement. The density around the side chain of serine is bit wide but the placement of its side chain closer to the carbonyl carbon of ligand was reverted back after refinement. The placement and fitting of other two (hydrolized and non-covalent) species seem to be optimum and no negative density was observed after refinement. These ligand species varies slightly in atomic composition due to addition or deletion of few atoms by virtue of acylation and deacylation reaction but the density at this resolution does not allow us to identify the exact species. There are four mol/ASU and the orientation of some of the atoms of ligand are different among them. I am very much eager to get some insights for few of my quarries:
1- How can I ascertain the that which of the ligand species is actually present in my structure? 2- If more than one species is present and observed density is the average of them then how to find out the proportion of each of the ligand species? 3- As mentioned earlier that I was able to fit all the forms of ligand with ocuupancy-1 in density without any negative density, why so? 4- How can we decide the occupancy factor for different species of same ligand? Based on assumption or is there any way to decide this? 5- If finally I could model the ligand in any of the form than how can we validate its correctness? 6- Is there any way to sharpen the density at lower sigma by excluding noise. Thank you. Shiv
