Eric, as always, you have such great ideas...
On Thu, Jul 28, 2011 at 5:05 PM, Eric Martz <[email protected]>wrote:
> Hi, Bob,
>
> 1. What do I mean by "break FirstGlance"? Just that the proposed changes
> may break a number of existing resources (e.g. Proteopedia). For FirstGlance
> and Proteopedia specifically, I don't mind at all tweaking its scripts to
> accommodate what you decide to do, Bob. That is no problem given that I am
> now alerted to watch out for this change. I agree that having a better
> definition of "ligand" in Jmol is desirable.
>
>
OK, that makes sense.
> 2. From RasMol and Chime and early Jmol, I long ago accepted that "ligand"
> was "hetero and not solvent" because that's how those programs have been
> structured. And their structure reflected limitations in pre-remediation PDB
> files. Therefore in FirstGlance I speak of "ligand+" and explain how it
> differs from the usual concept of "ligand".
>
>
I'm guessing, then, that you do not use the term "ligand" at all, which
would be good. Then I think there's nothing for you to fix, perhaps, but see
below....
> 3. I think any atom that is a member of a group named one of the 20
> standard amino acids or the 12 standard nucleotides should be "protein" or
> "nucleic" respectively. This would take care of the 3 cases of "anomalous
> atoms" I listed earlier, by deeming those atoms to be protein, and hence not
> "ligand".
>
>
Very nice. done. Jmol 12.2 only; See
http://chemapps.stolaf.edu/jmol/docs/examples-12/Jmol-12.zip
> 4. What would be an ideal definition of ligand? Bob, you are far better
> qualified than me to decide. Here are my thoughts:
>
> A. Should covalently bonded adducts (e.g. carbohydrate glycosylation of
> proteins) be excluded from "ligand"? Well, according to wikipedia, in
> inorganic chemistry, ligand refers to the binding of electron donor
> molecules to metal ions. Such binding has some covalent character. In
> biochemistry, one usually thinks of ligand binding as reversible -- but can
> we clearly distinguish ligand from substrate? And do we want to exclude
> covalently-linked moieties from "ligand"? I think I favor avoiding this
> problem by leaving covalently linked non-(protein,nucleic,water) within
> "ligand".
>
>
Here you go: @ligand !(protein,dna,rna,water,urea)"
> B. Should non-standard residues within protein and nucleic acid chains be
> excluded from ligand? I think yes. Easy since those are (almost always)
> HETATM.
>
>
nonstandard residues have always been considered protein or nucleic if they
have the required atoms (C-alpha for proteins, a few atoms for nucleic)
> There is at least one case where "protein and hetero" will fail to detect a
> non-standard residue: 1b07. In this case, the proposed new definition of
> "ligand" will still include this non-standard residue, but this situation is
> rare (possibly unique) so it is OK. (In 1b07, a non-standard residue is
> represented in the PDB file as ATOM GLY plus HETATM PYL. The PYL
> (phenylethane) is covalently linked to the nitrogen of a glycine, making an
> "amide N-substituted residue" that fools one receptor into thinking it is
> proline. There are 4 other PDB entries containing PYL: in 3, the PYL is a
> noncovalently bound true ligand; in the 4th (2vrm) PYL is covalently bound
> to the true ligand FAD.
>
>
not much we can do about PYL, I think. I'm happy to say, though, that
set pdbAddHydrogens; load =1b07
nicely turns that aromatic and adds the proper hydrogens. It's basically a
covalently bound ligand, like retinal in rhodopsin.
> C. Water is a true ligand but because of the quantity often present in PDB
> entries, it is convenient to separate it from "ligand".
>
> Water should include deuterium oxide (DOD as in 3qf6 and 55 other entries;
> not D2O [see 3nzc]). Jmol currently does this, since "water", "hoh" and
> "dod" all select the same 199 atoms in the fully deuterated model 3qf6.
>
>
with the latest fix, "HOH" will not find the D2O in 3qf6. I think that is
proper.
> However, non-water "solvent" (notably sulfate ions) should, in my opinion,
> be included in "ligand".
>
>
done.
> So in conclusion, I agree with Bob's suggestion but with the modification
> that "solvent" become "water", so the new definition of "ligand" would be
>
> *not (protein,nucleic,water)
>
> *
>
in conclusion, for Jmol 12.2 only:
@ligand !(protein,nucleic,water,urea)
"PO4" and "SO4" will be ions, which are a subset of ligand
"SOL" will be sorbitol, as it should be, a carbohydrate, which is a subset
of ligand
anomolous atoms - anything with a residue name matching the known amino
acids or nucleic acids will be treated appropriately
it is possible for a group to be considered nucleic and not DNA or RNA,
PURINE, or PYRIMIDINE, though, because Jmol can't determine if a nonstandard
nucleic acid is one of those by name alone. (I think....)
Jmol 12.2 only; See
http://chemapps.stolaf.edu/jmol/docs/examples-12/Jmol-12.zip
Bob
------------------------------------------------------------------------------
Got Input? Slashdot Needs You.
Take our quick survey online. Come on, we don't ask for help often.
Plus, you'll get a chance to win $100 to spend on ThinkGeek.
http://p.sf.net/sfu/slashdot-survey
_______________________________________________
Jmol-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jmol-users
