I am solving a crystal structure using molecular replacement in phenix. The mutated residue is Glycine to gultamate and present in a turn. While refining in the first round I selected rigid body refinement and 3 cycles. When observed in coot I can find electron density however as I add water in second round and refine for XYZ (reciprocal space, real space), ocupancies, indivdiual B factor (5 cycles), the electron density over the glutamate is lost. Resolution obtained is 2.2 angstrom and I/sigma is 3.
So what could be the probable reason for the same and how can tackle this issue?
Shubhashish Chakraborty What do you mean with electron density is lost? Did you change the glycine to glutamate in the model before doing the second refinement? Has the density been filled with water molecules maybe?
By the way, are you sure you want to truncate your data at an I/sigma of 3 ? Any reason why you are discarding weaker reflections ?
So the template I am using has glycine and the protein I am working with has glutamate at the same position (as its a mutant of the template protein).
I have changed the amino acid after phasing and dis the first phase of refining. Post that in the second phase I added water molecules and set the model for refining however I couldn't see the electron density now.
No, the density isn't filled by water molecules.
Sorry, but I couldn't get the second part of your question.
To the second point: Depending which type of detector and x-ray source you have used, many people would use waek reflections at much lower I/sigma (something between 1 and 1.5) assuming all other statistics (completeness, CC(1/2), ... ) are fine, and multiplicity is decently high. This could improve not only your resolution, but also your overall map quality.
Yet I am a bit puzzled by what you describe. What would happen if you don't change the residue in the model after phasing, do rigid body and "full" refinement (avoid updating the water molecules at this stage)? After that full refinement, you should see a clear Fo-Fc density for your side chain. Do you see this one already after MR? Can it be that the sidechain is partially flexible?
Christophe Wirth Yes I have changed the I/sigma value to 2 but there isn't much change in resolution as I can see after scaling.
Yes, the amino acid is present in a turn and can be dynamic. After phasing too I can't find electron density over glycine backbone.
So the best is to remove a few residues around your Gly/Glu from your model, do refinement without water update. If difference density appears, then you can rebuild. If you don't have difference density, the loop is likely to be flexible/disordered.
- Badges
- Science topic
- Similar topics
- Biological Science
- Biochemistry
- Enzymology
- Enzymes