NMR Restraints Grid

Result table
 (Save to zip file containing files for each block)

image mrblock_id pdb_id cing stage position type subtype
12463 1szy cing 1-original 20 protocol structure calculation


Distances:
    A-form NOEs for stem:
 cp ../stem/noes.inp noes.stem.inp
 Hand-edited noes.stem.inp to change segid b resid 1-7 to segid b resid
     15-21, then removed segids
    Real NOES for loop:
 I'm artificially constraining the stem to A-form and there are no
     exchangeable NOEs in the loop (or between stem and loop) so I need
     concern myself only with nonexchangeable NOEs involving at least
     one loop proton.
 egrep 'resid ([8-9]|1[0-4])' ~/nmr/italy2-30c.int/noes.d2o.inp >
     noes.loop.inp.old
 cp noes.loop.inp.old noes.loop.inp
 Hand-edited

Dihedrals:
    Hand-created dihedrals.d2x using values from
 - inspection of 31P-1H COSY (alpha-zeta)
 - inspection of aromatic-sugar region of NOESY (chi)
 - ~/nmr/italy2-30c.j/j2nu.out (nu0-nu4); left intermediate Js
   unconstrained and constrained U12 (75% C2'-endo) to C2'-endo
        - used A-form angles for G1 and C21, since I'm using A-form NOEs there
    dihedrals2xplor dihedrals.d2x > dihedrals.inp

Template:
    Ran run.make-template.
    Moved make-template.??.out, template.??.pdb and make-template.summary to
 template/.
    Examined templates; template/template.07.pdb has the lowest energy
 (out of a not very large range)
    Linked template/template.07.pdb to template.pdb.

Miscellaneous:
    Increased dgsa.inp's weighting of cdih term by 4x relative to the manual.
 Doing so eliminated almost all dihedral angle violations (although
 violations were usually only 5 degrees or so anyway) and did not
 increase (in fact may have reduced) NOE violations.
    Ran series of calculations varying times and timesteps in template.inp and
 dgsa.inp; see ../italy2.results/{template,dgsa}.
    Ran calculations including interesting NOEs predicted by 4tna; see
 ../italy2.results/33-36.
    Ran calculations omitting scary (not A-like, nonsequential) NOEs and then
 reincluding them; see ../italy2.results/scary.
    Ran calculations on each set of dihedral angles. The cheating sets always
        result in more tRNAPhe-like structures than the corresponding honest
 sets. The free and fixed sets come out about the same.
    Ran calculations with and without tcoupling=true in refine.inp (it was
 only omitted due to an error in the first place) and with an extra
 cooling step to get down to 0K. Neither made much difference; I'm now
 using the former (which I'd have used without thinking in the first
 place if I hadn't left it out by accident) but not the latter.
 See ../italy2.results/tcoup,cool
    Ran calculations with U12 ribose free; didn't make much difference.
    Increased NOE weighting by twofold relative to the manual. See
 ../italy2.results/noe-weighting

Averaging:
    Determined that 3, 5, 8, 11, 14, 17, 18, 20, 22, 23, 24, 27 had converged.
    Ran average-dihedrals on those structures. Observed that A11 alpha, beta
 and gamma are in one distinct range in 8, 18, 27 and another in the
 rest.
    Ran average-dihedrals on 3, 5, 8, 11, 14, 17, 18, 20, 22, 23, 24, 27 and
 on 8, 18, 27.
    Will report average across all converged structures for all angles except
 A11 alpha, beta, gamma, for which I'll report averages for each subset.
    Ran average.inp once on all 12 and once on all but 8, 18, 27. The
 averaged structure is the same. Threw away the former.


Please acknowledge these references in publications where the data from this site have been utilized.

Contact the webmaster for help, if required. Thursday, March 28, 2024 12:21:06 PM GMT (wattos1)