Hello everyone. Many thanks for the thoughts on the matter. Dear Dr Abhinav
You can try to run a simulated annealing in CNS and then try to refine that > model and see if that solves the problem. This would most likely help in > getting rid of the overfitting problems. > I have trouble installing CNS on my machine which has Windows OS. Nevertheless, rigid body refinement (RBR) and simulated annealing (SA) (and combinations thereof) with phenix led to distant Rf and Rw (a sign of model bias) at the beginning of refinement and even more so towards the end (*at tedium*). Toggling off of RBR and SA, however, bridged the gap between Rf and Rw while lowering them at the same time. Dear Dr Mark The R/Rfree values seem very low for a 2.67Å resolution structure. Perhaps > it was over-refined? Difficult to know without looking at the specific case > and details of the data quality though. > PFA merging stats. Asymmetric unit seems to have 8 molecules @~44% bulk solvent content. I can only speculate as to systematic problems arising in intensity correction accompanying the MR run on account of a combination of NCS (~26% off-origin peak in Patterson map) and a screw axis symmetry (P 1 21 1 being the space group). The best way to resolve validation concerns is to inspect the outliers > individually - it might be worth getting a second pair of experienced eyes > to look at them. Perhaps there is a peptide flip or a cis-peptide that you > missed. > Finally, it is not usually possible to remove all validation concerns, > especially at intermediate resolution like 2.67Å. > Among the clashes and the other geometry outliers, there are cis-prolines in the final model - same as the MR model that was pulled from RCSB. As such, I decided to leave them in. Trying to fix them in coot serves no good, as they come undone upon uploading the resultant structure to wwPDB. Over-fitting, as you pointed out, was also a result of the umpteen attempts to iron these issues out with refinement/remodelling. Dear Professor Eleanor I use the validation report to check for obvious modelling errors but if > you can’t find any, you have to just submit the results of your > experiment... > I have sent a message to wwPDB curator requesting to admit the current model - with outliers. Response awaited. Dear Dr Powell I will submit the data to PDB-REDO Dear Professor Cooper Many outliers have been done away with. The ones that remain seem truly intractable. We have been trying to submit the structure in its current state barring only the red exclamation point on the validation reports section, even with green ticks on every other section. Thank you all for the helpful thoughts, again. Please revert for further information. Yours faithfully Nitin Kulhar On Mon, Sep 25, 2023 at 5:44 PM Jon Cooper <jon.b.coo...@protonmail.com> wrote: > Hello, unless it's changed over the last couple of years, the pdb rightly > used to let you deposit a structure with outliers, hopefully not too many, > though ;-0 > > I take it the stalling you reported probably comes from whoever is looking > over your shoulder. > > Best wishes, Jon Cooper. jon.b.coo...@protonmail.com > > Sent from Proton Mail mobile > > > > -------- Original Message -------- > On 23 Sept 2023, 02:29, Nitin Kulhar < > 00009dfccc771c91-dmarc-requ...@jiscmail.ac.uk> wrote: > > > Dear all > > We have refined (Refmac5) a crystallographic structure with Rw/Rf values > 0.19/0.22 (Resln 2.67). However, the deposition has stalled on account of > the wwPDB's preliminary validation report, which indicates map/model and > geometry issues, with each criterion containing a few instances. We tried > to correct these by varying overall geometry restraint weights, e.g. > decreasing overall weights from default value of 1.0, incementally > decreasing sigmas corresponding to the planarity restraint term. This did > not resolve the issues. > > In another approach, real space refinement by hand (against 2fo-fc) in > coot brought the geometry parameters within acceptable limits in addition > to improved apparent agreement with electron density (2fo-fc, sigma=1), but > uploading the resultant coordinates seems to undo the changes made in coot, > as indicated by reappearance of same outliers in the subsequent validation > report. > > I request your kind suggestions in this regard. Please also revert for any > further information. > > Thanks > Nitin Kulhar > PhD student > c/o Dr Rajakumara Eerappa > Macromolecular Structural Biology Group > Department of Biotechnology > Indian Institute of Technology Hyderabad > Kandi, Sangareddy > Telangana, India 502284 > > Disclaimer:- This footer text is to convey that this email is sent by one > of the users of IITH. So, do not mark it as SPAM. > > ------------------------------ > > To unsubscribe from the CCP4BB list, click the following link: > https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 > > -- Disclaimer:- This footer text is to convey that this email is sent by one of the users of IITH. So, do not mark it as SPAM. ######################################################################## To unsubscribe from the CCP4BB list, click the following link: https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCP4BB&A=1 This message was issued to members of www.jiscmail.ac.uk/CCP4BB, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
