Ok. I ll try to do it definitely, I don't try to avoid it. But I am tryong to get some data relatively fast, let s say within a week or two. It's not for publication etc. it's just for guidlines of experiment (but that is the very rigid requirement - for modeling to predict experiment).
I plan of course to use QM approach for example as you suggested using your server probably, but it will take time anyway.
For now then I'll try to play with constraining part of the system etc.
Probably I can try to impose some not very strict constraints to the whole protein, except small molecule ligand which i am studying. Or I'll remain just 10-15 amino acids unconstrained and unconstrained ligand.
The main purpose is to understand whether the ligand binds supposed binding site. Docking gives too much false positives to use this method as reliable for such a purpose.
Best regards,
Andrew
18.01.10, 12:41, "FyD" <fyd.q4md-forcefieldtools.org>:
> Dear Andrew,
>
> Your Zinc atom is bound to amino-acids. Consequently, treating Zinc as
> a +2 ion is quite simple but might not lead to realistic conclusions.
> Deriving the charge value of a metal embedded in a bio-inorganic
> complex results in a quite important decrease of the charge value in
> comparison to that of the corresponding isolated ion.
>
> If you use a 'too large' charge value for your metal embedded in a
> bio-inorganic complex and/or if you constrain your system (using the
> metal ion value), your MD system might blow up & crash (observed
> problem).
>
> Thus, if you do a "dirty & cheap" parametrization, you might even not
> be able to get a MD trajectory. Consequently, I think you should pay
> the price of a rigorous parametrization for your metal complex.
>
> This is just my personal opinion based on MD simulations...
>
> regards, Francois
>
>
> > The parametrization goes ok. I get no errors. But as far as in
> > protein Zn is not an ion and it is bound to four surrounding amino
> > acids then of course when I parametrize zinc as ion it losts binding
> > to amino acids and the structure starts to become corrupted. So I
> > think that I definitely just need to fix ainc atom and these four
> > amino acids. This part of the structure should be fixed, probably
> > even more amino acids because they are far from the binding site,
> > but can innfluence the whole geometry and the binding site. Can I
> > use it with parameters as below? Probably I should use just the part
> > of the protein around the binding site and fix everything else?
> >
> > Best regards,
> > Andrew
> >
> > 2rf5
> > $cntrl
> > imin = 0, irest = 1, ntx = 7,
> > ntb = 2, pres0 = 1.0, ntp = 1,
> > taup = 2.0,
> > ntc = 2, ntf = 2,
> > tempi = 300.0, temp0 = 300.0,
> > ntt = 3, gamma_ln = 1.0,
> > nstlim = 2500000, dt = 0.002,
> > ntpr = 100, ntwx = 5000, ntwr = 1000
> > /
> > Keep protein fixed with weak restraints
> > 10.0
> > RES 131, 134, 139, 142, 208
> > END
> > END
> >
> > 13.01.10, 09:47, "Ross Walker" :
> >
> >> Hi Andrew,
> >>
> >> If you think the zinc is purely structural and not involved in the actual
> >> binding site then you can just model it as a 2+ ion. Put TER cards
> >> around it
> >> in the pdb. Call it something like residue name ZNS (for structural Zinc)
> >> and give it atom name ZN.
> >>
> >> Then fire up leap and do:
> >>
> >> edit ZNS
> >>
> >> This will create a new unit called ZNS.
> >>
> >> Draw in a single new atom and then highlight and edit it.
> >>
> >> Set the charge to +2, the name and type to ZN.
> >>
> >> Then
> >>
> >> savemol2 ZNS ZNS.mol2
> >>
> >> Quit Leap.
> >>
> >> Create an frcmod file with the contents:
> >>
> >> Zinc 2+ Params
> >> MASS
> >> Zn 65.38
> >>
> >> BOND
> >>
> >> ANGL
> >>
> >> DIHE
> >>
> >> NONB
> >> Zn 1.85 0.06
> >>
> >>
> >> Finally you can fire up leap:
> >>
> >> source leaprc.ff99SB
> >> ZNS = loadmol2 ZNS.mol2
> >> loadamberparams frcmod
> >> foo = loadpdb foo.pdb
> >>
> >> And your zinc should be recognized.
> >>
> >> The only other thing you might have to do is make sure the
> >> protonation state
> >> (HIS/HID/HIP, CYS/CYX etc) is correct for the residues surrounding the zinc
> >> so you don't get H's added very close to the zinc.
> >>
> >> Good luck,
> >> Ross
> >>
> >> > -----Original Message-----
> >> > From: amber-bounces.ambermd.org [mailto:amber-bounces.ambermd.org] On
> >> > Behalf Of Andrew Voronkov
> >> > Sent: Wednesday, January 13, 2010 2:11 AM
> >> > To: AMBER Mailing List
> >> > Subject: [AMBER] How to avoid Zn parametrization?
> >> >
> >> > Dear Amber users,
> >> > I have a protein which has Zn atom, bound directly to four amino acids
> >> > (Tankyrase PARP domain, 2rf5 code in PDB bank).
> >> > I need to test the results of docking of small ligands by MD run by
> >> > evaluation of the stability of complexes. The binding site is far from
> >> > Zn atom. For now I have no time to make parametrization for Zn atom as
> >> > far as it seems to be rather advanced and time consuming. What options
> >> > do I have?
> >> > For example I haven't included Zinc in docking site while performing
> >> > docking. Can I maybe make CAP molecular dynamics only for the binding
> >> > site of the ligand.
> >> > Probably parametrize zinc as ion, or just cut out Zn atom and fix all
> >> > Zn-surrounding amino acids as in the X-ray structure?
> >> >
> >> >
> >> > Best regards,
> >> > Andrew
>
>
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Mon Jan 18 2010 - 04:30:03 PST
