-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
In general the word "outlier" should not be read as "error". It simply means that this thing is interesting. You have to judge, based on all information at hand, if it is interesting because it is rare or because it is wrong. It is true that the rmsd of high resolution models tend to be higher than those of low resolution. The libraries we use are distillations of high resolution structures and do not contain all of the true variability of reality. We have to use libraries when the diffraction data is insufficient for the model to stand on its own, but you have to keep in mind the library's limitations. Dale Tronrud On 9/15/2014 8:42 AM, C wrote: > Isn't Molprobity too tight anyway for _high resolution_ structures? > > > I have often (for proteins) found it highlighting "outliers" when > in fact that is what the density shows. > > Isn't it a trade off between molprobity and what the measurements > show, especially at high resolutions? > > > -----Original Message----- *From:* merr...@u.washington.edu *Sent:* > Mon, 15 Sep 2014 08:35:37 -0700 *To:* ccp4bb@jiscmail.ac.uk > *Subject:* Re: [ccp4bb] Bond lengths and angles used by Molprobity > for ANP (AMPPNP) > > On Monday, 15 September 2014 03:24:19 PM Andrew Leslie wrote: > >> Does anyone know if Molprobity has recently changed the standard > bond lengths and angles that it uses ? > >> > >> Molprobity is reporting errors in the C4-C5 bond length and the > C5-C4-N3 bond angle (deviations of 8-10 sigma) for AMPPNP (monomer > code ANP) for a new structure refined with Refmac. I then tried > Molprobity on a deposited PDB structure that also contains AMPPNP > and it reported the same errors. I am sure that these errors were > not reported when this structure (2JDI) was deposited in 2007. > >> > >> This would suggest that the standard dictionary that Molprobity > uses has changed, but I cannot find any reference to this on the > Molprobity pages. > >> > >> I would be very grateful if anyone can throw some light on this. > > > > I suspect, though I cannot say for sure, that the difference is > > whether ANP is [correctly] treated as a ligand or [incorrectly] > > treated as a monomer in a nucleic acid polymer. We have seen > > similar cases recently. For now I would say it is safe to > > disregard Molprobity scores for ligands that just happen to be > > nucleic acids. > > > > Ethan > > > > > > > > > >> Thanks, > >> > >> Andrew > > > > -- > > mail: Biomolecular Structure Center, K-428 Health Sciences Bldg > > MS 357742, University of Washington, Seattle 98195-7742 > > ------------------------------------------------------------------------ > > Can't remember your password? Do you need a strong and secure password? > Use Password manager! It stores your passwords & protects your > account. <http://mysecurelogon.com/password-manager> -----BEGIN PGP SIGNATURE----- Version: GnuPG v2.0.22 (MingW32) iEYEARECAAYFAlQXCzoACgkQU5C0gGfAG13CJACcDqiw/U+UVU7+L8MJi1i5u0mp NZwAnAzQTFAFStnpLuIcwfhgPK81F6aD =gh5L -----END PGP SIGNATURE-----
