[Chimera-users] mol2 for NH3
Elaine Meng
meng at cgl.ucsf.edu
Thu Jun 11 15:45:31 PDT 2015
As mentioned previously, there is a high incidence of charge-calculation failure for such highly charged molecules, and as you saw, the AM1 method depends on the coordinates of the atoms (the 3D structure), not just the bond connectivity like Gasteiger. Thus even the same exact molecule in different conformations could have different results.
Thus, no use in asking me whether it will work on your specific molecule… you’d have to try and find out. I don’t know the explanation.
Also, keep in mind that Antechamber is a separate program from Chimera (developed by a separate group, although included with and used by Chimera) and that it is not mean to handle all possible molecules. From <http://www.rbvi.ucsf.edu/chimera/docs/ContributedSoftware/addcharge/addcharge.html>:
"Note that Antechamber/GAFF are meant to handle most small organic molecules, but not metal complexes, inorganic compounds, or unstable species such as radicals, and may not work well on highly charged molecules."
I would not say that ignoring the guess is the reason why Gasteiger works. Gasteiger is just simpler than the other method (more “quick-n-dirty”). Maybe you could conclude it was at least partly the reason if AM1 works when you change the guess to -4 from -3, but I don’t know if you did that.
Best,
Elaine
> On Jun 11, 2015, at 3:31 PM, MPI <mpi566 at gmail.com> wrote:
>
> Hi Elaine,
>
> Thank you very much for the explanations. It is clear to me why
> gas method works by ignoring the calculation of net charge in Chimera.
>
> I also noticed that Eric mentioned ligand preparation for FAD or NAD
> and provided scripts eg, dockprep.py through DockPrep in Chimera to
> generate mol2 but it fails for some cases of drugs/small molecules
> when it uses am1 method and gets choked with sqm.
>
> Can you explain two examples as it fails for NADP from PDB 1w6u
> but succeeds for NAD from PDB 2h4f ? Of course, it works well for
> both, using gas method.
>
> Can a mol2 file of NADP be generated "with am1 method" by Chimera ?
>
> Thanks,
> Dewey
>
> # Here are the NADP coordinates prepared from PDB 1w6u.
> HETATM 8621 O4B NAP A1329 27.294 6.186 24.430 1.00 31.96 O
> HETATM 8622 C3B NAP A1329 29.273 7.323 24.965 1.00 31.96 C
> HETATM 8623 O3B NAP A1329 29.117 8.708 24.774 1.00 31.54 O
> HETATM 8624 C2B NAP A1329 28.635 6.867 26.257 1.00 31.84 C
> HETATM 8625 O2B NAP A1329 28.669 7.864 27.246 1.00 32.77 O
> HETATM 8626 C1B NAP A1329 27.211 6.566 25.797 1.00 31.30 C
> HETATM 8627 N9A NAP A1329 26.569 5.508 26.595 1.00 29.72 N
> HETATM 8628 C8A NAP A1329 26.965 4.203 26.733 1.00 29.33 C
> HETATM 8629 N7A NAP A1329 26.094 3.560 27.542 1.00 29.19 N
> HETATM 8630 C5A NAP A1329 25.141 4.438 27.924 1.00 29.33 C
> HETATM 8631 C6A NAP A1329 24.024 4.317 28.742 1.00 29.04 C
> HETATM 8632 N6A NAP A1329 23.810 3.204 29.446 1.00 28.84 N
> HETATM 8633 N1A NAP A1329 23.221 5.420 28.943 1.00 27.78 N
> HETATM 8634 C2A NAP A1329 23.512 6.631 28.345 1.00 28.64 C
> HETATM 8635 N3A NAP A1329 24.622 6.744 27.534 1.00 29.91 N
> HETATM 8636 C4A NAP A1329 25.422 5.665 27.329 1.00 30.00 C
> HETATM 8637 O1N NAP A1329 31.899 5.159 18.612 1.00 31.81 O
> HETATM 8638 O2N NAP A1329 30.918 7.458 19.154 1.00 31.66 O
> HETATM 8639 O5D NAP A1329 29.461 5.537 18.415 1.00 31.27 O
> HETATM 8640 C5D NAP A1329 28.309 6.346 18.394 1.00 29.19 C
> HETATM 8641 C4D NAP A1329 27.184 5.571 17.700 1.00 28.26 C
> HETATM 8642 O4D NAP A1329 27.552 5.194 16.382 1.00 27.30 O
> HETATM 8643 C3D NAP A1329 26.819 4.284 18.421 1.00 28.15 C
> HETATM 8644 O3D NAP A1329 25.416 4.141 18.367 1.00 27.49 O
> HETATM 8645 C2D NAP A1329 27.461 3.193 17.600 1.00 27.46 C
> HETATM 8646 O2D NAP A1329 26.705 2.021 17.677 1.00 27.27 O
> HETATM 8647 C1D NAP A1329 27.438 3.801 16.202 1.00 27.15 C
> HETATM 8648 N1N NAP A1329 28.517 3.306 15.333 1.00 27.66 N
> HETATM 8649 C2N NAP A1329 29.780 3.831 15.412 1.00 28.55 C
> HETATM 8650 C3N NAP A1329 30.781 3.348 14.576 1.00 30.27 C
> HETATM 8651 C7N NAP A1329 32.172 3.917 14.648 1.00 31.70 C
> HETATM 8652 O7N NAP A1329 33.086 3.524 13.659 1.00 32.49 O
> HETATM 8653 N7N NAP A1329 32.510 4.759 15.629 1.00 31.91 N
> HETATM 8654 C4N NAP A1329 30.499 2.342 13.655 1.00 30.48 C
> HETATM 8655 C5N NAP A1329 29.207 1.820 13.569 1.00 29.50 C
> HETATM 8656 C6N NAP A1329 28.227 2.325 14.423 1.00 29.09 C
> HETATM 8657 P2B NAP A1329 28.809 7.452 28.801 1.00 33.99 P
> HETATM 8658 O1X NAP A1329 29.948 6.524 28.942 1.00 33.70 O
> HETATM 8659 O2X NAP A1329 29.037 8.749 29.590 1.00 34.45 O
> HETATM 8660 O3X NAP A1329 27.518 6.755 29.259 1.00 32.91 O
> HETATM 8661 PA ANAP A1329 31.539 5.835 21.912 0.50 35.79 P
> HETATM 8662 O1AANAP A1329 32.626 4.830 21.987 0.50 36.39 O
> HETATM 8663 O2AANAP A1329 32.125 7.258 21.784 0.50 34.61 O
> HETATM 8664 O5BANAP A1329 30.638 5.689 23.240 0.50 34.77 O
> HETATM 8665 C5BANAP A1329 29.241 5.476 23.204 0.50 33.46 C
> HETATM 8666 C4BANAP A1329 28.496 6.644 23.846 0.50 32.43 C
> HETATM 8667 O3 ANAP A1329 30.551 5.483 20.682 0.50 34.18 O
> HETATM 8668 PN ANAP A1329 30.815 5.984 19.171 0.50 31.74 P
>
>
>
>
> On 6/11/15, Elaine Meng <meng at cgl.ucsf.edu> wrote:
>> Hi Dewey,
>> The guessing is relatively simple and yes, it does set Chimera atom types so
>> that adding hydrogens will generate protonation states reasonable at pH near
>> 7, as described here:
>> <http://www.rbvi.ucsf.edu/chimera/docs/ContributedSoftware/addh/addh.html#states>
>>
>> However, Chimera does not do any sophisticated pKa predictions, it just
>> looks at the functional group itself and tries for the most likely state
>> (for example, negative carboxylates, positive primary amines). If you are
>> highly concerned with the protonation states, or are working with ligands
>> that might have multiple states, you might want to consider getting your
>> ligand structures from some database like ZINC that has multiple protonation
>> states and allows specifying different pHs instead of doing it all in
>> Chimera from scratch.
>>
>> NAP is NADP. We have already noted that such highly charged molecules are
>> more likely to have problems in addcharge, and in fact we just use lookup
>> tables for ATP, ADP, GTP, and GDP:
>> <http://www.rbvi.ucsf.edu/chimera/docs/ContributedSoftware/addcharge/addcharge.html#cofactors>
>> …taken from this database, which also has the various states of NAD and
>> NADP:
>> <http://www.pharmacy.manchester.ac.uk/bryce/amber/>
>>
>> You could decide which NADP you want for your docking (or use more than
>> one), and just use the charges from that database. Although you could apply
>> them in Chimera, that would probably involve a lot of manual editing, so you
>> might just want to manually edit them into your output Mol2 file directly
>> instead of trying to get them in Chimera and then writing out the Mol2
>> file. How to apply custom charges in Chimera is described in the last
>> section of the Add Charge man page, “reading charges from files”:
>> <http://www.rbvi.ucsf.edu/chimera/docs/ContributedSoftware/addcharge/addcharge.html#input>
>>
>> The warning message is because the Gasteiger method actually ignores the
>> Chimera net charge guess, as per:
>> <http://plato.cgl.ucsf.edu/pipermail/chimera-users/2012-March/007336.html>
>>
>> So the charges you get add up to -4, not -3, which might be the “right”
>> answer anyway.
>> I hope this clarifies,
>> Elaine
>> ----------
>> Elaine C. Meng, Ph.D.
>> UCSF Computer Graphics Lab (Chimera team) and Babbitt Lab
>> Department of Pharmaceutical Chemistry
>> University of California, San Francisco
>>
>>> On Jun 11, 2015, at 11:58 AM, MPI <mpi566 at gmail.com> wrote:
>>>
>>> H Elanine,
>>>
>>> Thank you for the detailed suggestions. It works.
>>>
>>>> Chimera guessed the most likely charge state as and assigned the nitrogen
>>>> the atom type of N3+, which would result in 4 hydrogens added and net
>>>> positive charge.
>>>
>>> I am concerned about the style of "Chimera guessed" because many
>>> small molecules and drugs need to be screened.
>>>
>>> When a small molecule is assigned with partial charges by the
>>> command addcharge with am1 method, I noticed that multiple cases
>>> failed. With gas method, some cases will pass. However, Chimera gave
>>> a warning of unequal to the total formal charge (see below).
>>>
>>> What do you think about this ? Is it related to idatmType ? If so, it
>>> seems to me that one has to exam each atom of a small molecule if the
>>> process fails, especially with am1 method.
>>>
>>> Take NAP as example.
>>>
>>> eg, a warning of net charge
>>>
>>> Adding standard charges
>>> Charge model: AMBER ff14SB
>>>
>>> Standard charges added
>>> Assigning partial charges to residue NAP (net charge -3) with gas method
>>> ....
>>> ....
>>> (NAP) does not equal to the total formal charge (-4.00)
>>> according to the Gasteiger atom type
>>>
>>> Thanks,
>>> Dewey
>>
>>
>
More information about the Chimera-users
mailing list