To refine the structure, I use the Phenix and Coot programs. I have already removed the loop, rebuilt it using Coot algorithms and try to fit residues manually, the side chains of loop still do not want to fit into the electron density. Maybe it's better to do it in other programs or you know some tutorials about it. Thanks.
Dear Sabina Diusenova ,
if you are using COOT and PHENIX correctly, you are using some of the most used and robust programs available in the X-ray crystallography field.
In my view, there two hypothesis: either you do not have enough density to fit this loop (a printscreen would help visualising this), or you are misusing/misfitting in COOT/PHENIX.
Other details would also be helpfull: resolution, data processing statistics (Rmerge, Rpim, completeness and CChalf), stage of crystallographic refinement (are you in the beggining or at late stages? are there any other loops, ligands, solvent to model, or just this loop?), etc.
If you want, I can try to help off Research Gate in order not to make your data public. Please get back to me if you think I could be of assistance (my e-mail is [email protected]).
Best regards,
Jose
Dear Jose A. Brito ,
I sort of managed to refine this loop. Mostly I was confused by the fact that I did not see the side chain of tyrosine (11Tyr) and there was a rather large area next to the loop where there was not enough electron density (the green one). You can see from the photos that, in general, nothing has changed much, the green area has become a little smaller, and tyrosine is still not visible. Other residues fit quite well into electron density, except for tyrosine, so it may not be tyrosine at all. I have already asked the people who synthesized this protein about it, they should answer me soon.
I thought maybe there are any other software for refining loops, I saw articles about using AI for loops modelling .
In general, I have quite good data: Rmerge = 9.5, Rmeas = 11.4, CC1/2 = 99.0 and completness = 96.6%. And quality of model: Rwork = 19.83, Rfree = 26.18%. I still have problems with rotamers outliers and adding water molecules, but I hope I can figure out by myself.
Anyway thank you for the offer of help :)
Kind regards,
Sabina Diusenova.
Dear Sabina,
congrats on solving the issue.
Just a few comments for your next moves:
1. SLOOPS (in CCP4) and LOOPY (in ARP/WARP), are good programs for you to try automated model building of those loops; details here: https://strucbio.biologie.uni-konstanz.de/ccp4wiki/index.php/Loop_modelling
2. errors in residue's sequence are quite common; do not believe a sequence, unless you have a sequenced plasmid or a mass-spec analyzed protein; so, TYR might not be TYR;
3. a difference of almost 7% might indicate overfitting; if you solved you structure by MR, take in consideration model bias and consider simulated annealing to "heat up" your structure and validate your MR solution;
4. besides REFMAC and PHENIX, BUSTER-TNT is also amazin, especially the "-L" flag to improve poor electron density.
HTH, best regards,
Jose
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts