Thanks to all who responded to my inquiry: Do carbon scattering
include a component for hydrogens?
Answer: no. Not a good idea. Yet our biological structures include hydrogens since
1/2 the atoms in a protein are H. We don't include them at low resolution so we can
'do no harm' to the refinement. They only contribute scattering at low resolution yet this is
the data that is present in low resolution structures (although hard to measure
accurately). So lets include them in all refinements, regardless of resolution. At least we
will have given the community a model with the most correct biological representation.
If necessary, lets report an Rfree with and without hydrogens.
And at what point do we 'see' the hydrogen? 2.5 sigma difference map? and what does it
look like? a 'bump', a 'knob', and with an equal number of H atoms as heavy atoms, a
typical 325 AA protein contains 3250 hydrogen atom. How much time will it take to
check all these hydrogens? If we only select those that we
see, how bizarre this will look in pymol.
So we should have all structures deposited with H refinement, either riding or
explicit (if seen). Then we can also better understand the oxidation states of ligands,
cofactors, etc by looking at the exchangeable hydrogens. And we should be able to
now add restraints for hydrogen bonding in alpha helices and beta sheets since we know
these are stable.
thanks
kas
--
Kenneth A. Satyshur, M.S.,Ph.D.
Associate Scientist
University of Wisconsin
Madison, Wisconsin 53706
608-215-5207
|