Hello John--

> I've have some RDCs with low SNR and respectively high uncertainty. 
> On the RDC page of the XPLOR manual, I see the RDC energy function is E = K
> (exp-calc)^2 with no uncertainty component. 
> Is this actually an abbreviated form, with the real one being E = 
> K((exp-calc)^2)
> /(Unc^2)?
> 

Some description of RDC weighting is given here:

  C.D. Schwieters, G.A. Bermejo and G.M. Clore, ``Xplor-NIH
  for molecular structure determination from NMR and other
  data sources,'' Protein Science, 1688, 311-240 (2018).
  https://doi.org/10.1007/978-1-4939-7386-6_14

The default RDCPot term does not currently have an option for
per-restraint weighting. Instead, restraints with different signal to
noise are placed in separate energy terms which are then scaled
appropriated. This is how experiments involving different nuclei are
handled in the script eginput/gb1_rdc/refine.py in the Xplor-NIH
distribution. 


> I also see in the Sani subsection the line: "POTEntial <SQUAre | HARMonic> !
> whether to use err or not."
> 

Square potType is typically only used for through-space RDCs, between
atoms which are not bound. 

best regards--
Charles

_______________________________________________
Xplor-nih mailing list
[email protected]
https://dcb.cit.nih.gov/mailman/listinfo/xplor-nih

Reply via email to