On Sat, 2012-05-12 at 19:28 +0100, Yuri Pompeu wrote:
> Dear community,
> I am probably disturbing a sleeping bear
definitely so
> Reading the thread on hydrogen deposition with the model, I came accross several arguments that make sense on their own, but when put together are puzzling and dont seem to converge to an answer.
this is true for most other recurring threads
> -Some argued that depositing riding hydrogens with the model may imply that your data had enough information for you to include hydrogen atoms in the final model.
Unless there is a remark that explicitly states that these are riding
hydrogens, but nobody reads those
> This is definetely a problem especially when dealing with non-experienced users that may think the model is more accurate than it really is.
they always do and they have no idea how accurate the model actually is
> -It seemed to be consensus, that when softwares use hydrogen restraints it can be beneficial geometrically and also can make your model a better description of your x-ray data.
I think nobody disputes that, although the benefit may vary from
structure to structure
> Based on these two main arguments, many would agree that hydrogens should be included throughout refinement but not deposited.
I do agree and I won't deposit them myself, but then what others choose
to deposit is really their choice.
> So this brings me to last point that was also mentioned in the old thread. If you used riding hydrogens throughout refinement and arrived at a final model that you believe best describes your x-ray data to a certain level of accuracy (Rvalues, geometry, map CC, etc...) would you not be invalidating the whole refinement process by going in and removing the hydrogen atoms right before deposition?
Not really. You report that you used riding hydrogens and you report
the program you used to generate them. In theory, anyone can dig up the
appropriate version and reproduce your results.
> So how would one avoid this Catch-22?
I don't think it's strictly a catch-22 situation. The issue is that
depending on what the structural model is used for, different forms of
the pdb file may become most useful. The only situation I can imagine
when having riding hydrogens is beneficial is for algorithm development
and perhaps verification of how much differences in riding hydrogen
treatment contribute to differences in things like R-values. Both are
quite esoteric tasks and you already provide sufficient information
(vide supra).
Cheers,
Ed.
--
"Hurry up before we all come back to our senses!"
Julian, King of Lemurs
|