Dear Ed,
> On 14 Nov 2015, at 21:40, Ed Pozharski <[log in to unmask]> wrote:
>
> Tassos,
>
> I agree that there is very high probability that disordered side chains
> are just that, disordered, and not missing from covalent structure. But
> there is also often very conclusive evidence that few more N-terminal
> residues are there and very conclusive evidence that certain loops are
> there. Should I model them too even though I don't know where exactly
> they are? Probably not.
I would say probably yes, if you could have a good guess.
> I just wonder why do you feel stronger (if you
> do) about side chains than about 3-residue loop, even though there is
> roughly the same number of atoms involved and you can place them with
> roughly the same coordinate uncertainty.
The side chain has a limited conformational space that it can occupy.
That space can be represented pretty well by a single conformation - or better by multiple conformations.
In contrast, the conformational space occupied by ten N-terminal residues is massive.
We dont have a good way representing it.
>
> Sometimes there might be very conclusive biochemical evidence that
> certain non-covalent interactions must be occurring - e.g. a peptide
> derived from the antigen simply must bind in the well known site. But if
> electron density is not there - can I still model the peptide? This
> sounds like a slippery slope argument (because it is), and so I would
> certainly agree that comparing disordered side chains with imaginary
> peptides is ridiculous. And yet it reveals that the distinction is
> simply quantitative.
>
> Also, there is some evidence that removing disordered side chains from a
> model lowers Rfree.
Well, see above.
As you are pointing out there are good arguments on both sides, thats why
there is no clear conclusion yet !
It looks a 65% for B factors for now at the poll!
By Western parliamentary standards the issue would had been concluded ...
But, like in democracy, issues are not as simple as a vote ;-)
A.
>
> I apologize to everyone if I have unwittingly contributed to reanimating
> the dreaded ghost of side chain disorder discussion.
>
> Cheers,
>
> Ed.
>
>> There is very conclusive scientific evidence that a certain side chain
>> is there, from sequencing.
>> Its not from the same experiment, but its scientific and very
>> deterministic, there is no doubt of what is “there”.
>> We just don’t know where exactly it is ;-)
>>
>> A.
>>
>>
>
|