Great discussion !
So basically if I have a pdb file with all atoms (including hydrogens) but
I am not chemist at all (sorry for that !) so for me it's quite difficult
to say whether it would be positive, negative or totally neutral, do we
have some utility which may return to me the net charge of the molecule
based on its atomic composition ?
Yours with thanks
Enrico
Il giorno gio 16 feb 2023 alle ore 08:50 Steinbrecher, Thomas <
thomas.steinbrecher.roche.com> ha scritto:
> Hi Vinicius,
>
> sure, if all the hydrogens are set, the charge should be clearly defined.
> Even without that, as long as you have e.g. a SMILES with correct bond
> orders, the total charge should be trivial to compute (e.g. RDKit has a
> getFormalCharge(molecule) method which I have no experience with but which
> should do exactly that). Many more modelling tools will compute a total
> molecular net charge for you, it is one of the most common properties
> available.
>
> So we may need to split the question into multiple parts maybe:
>
> - Computing the total charge of one molecule is easy, either manually or
> with many different computational tools
>
> - The antechamber program specifically has, afaik, no way to do this
> automatically, because there usually is no need, the user setting up an MD
> will have a clear idea of what the charge state of their molecules are
> (hopefully...) and provide it with the -nc parameter.
>
> - If you are talking about million compound libraries, then you need to
> pick some ligand preparation tool that will assign net charges at the pH
> you want to study (and 3D coordinates and tautomers and bond order, and
> etc...) for all molecules in an efficient manner. The one I am most
> familiar with, ligprep from Schrodinger, handles this quickly even for
> large libraries by relying on common structure patterns (e.g. amines tend
> to be positive, carboxylic acids negative...) and other docking workflows
> will have similar tools. There it is important to remember that this
> assignment will not be perfect, there will always be some charge states and
> tautomers missed or wrong.
>
> Kind Regards,
>
> Thomas
>
>
>
>
> On Wed, Feb 15, 2023 at 9:44 PM Vinicius Wilian Cruzeiro <
> vwcruz.stanford.edu> wrote:
>
>> Hello Enrico and Thomas,
>>
>> I am chiming in because I am interested in this discussion 😉
>>
>> What Thomas said makes sense, but what if your structure has all the
>> hydrogens set? Then it should be possible to predict the net charge, right?
>>
>> For example, I think this problem appears in docking codes, right? Do
>> these codes predict what the total charge of a given ligand is (based on
>> the structure) or is the user required to specify it? Given that a ligand
>> library could have 1M+ compounds, options 1 seems more reasonable.
>>
>> I am curious to see what you and others have to say about that.
>>
>> Thank you!
>>
>> *VinÃcius Wilian D. Cruzeiro, PhD*
>> Postdoctoral Researcher
>> Department of Chemistry
>> Stanford University
>> Twitter: @vwcruzeiro
>> ------------------------------
>> *From:* Steinbrecher, Thomas via AMBER <amber.ambermd.org>
>> *Sent:* Wednesday, February 15, 2023 9:19 AM
>> *To:* Enrico Martinez <jmsstarlight.gmail.com>; AMBER Mailing List <
>> amber.ambermd.org>
>> *Subject:* Re: [AMBER] [Sender Not Verified] Net charge of ligand
>>
>> Hi Enrico,
>>
>> I dont think there is a general answer to that, the net charge of a ligand
>> is something the person preparing a system in Amber needs to make a
>> decision about. PDB files usually dont contain hydrogens, so the charge
>> state of a ligand can be ambiguous just based on an X-ray structure.
>> Usually, you can make a pretty good guess based on chemical intuition, the
>> ligand pose and the interactions made with the protein, but that will not
>> resolve every case. Simulating a molecule in several charge states to
>> decide which one matches experimental information best is a possible
>> solution.
>>
>> Kind Regards,
>>
>> Thomas
>>
>> On Wed, Feb 15, 2023 at 5:05 PM Enrico Martinez via AMBER <
>> amber.ambermd.org>
>> wrote:
>>
>> > Dear Amber users!
>> >
>> > Dealing with the preparation of the protein-ligand system, often I could
>> > observe errors in the Antechamber parametrization related to the NET
>> charge
>> > of the ligand.
>> >
>> > How could I determine precisely what the NET charge of the ligand based
>> on
>> > its atomic coordinates (e.g. providing PDB contained all atoms of the
>> > ligand as the input) ?
>> >
>> > Many thanks for advance
>> >
>> > Enrico
>> > _______________________________________________
>> > AMBER mailing list
>> > AMBER.ambermd.org
>> > http://lists.ambermd.org/mailman/listinfo/amber
>> >
>>
>>
>> --
>> *Dr. Thomas Steinbrecher*
>> Principal Scientist CADD
>>
>> Roche Pharma Research and Early Development
>> Roche Innovation Center Basel
>> F. Hoffmann-La Roche Ltd
>> Bldg. 092/3.92
>> Grenzacherstrasse 124
>> 4070 Basel
>> Switzerland
>>
>> Phone +41 61 682 1319
>> mailto: thomas.steinbrecher.roche.com
>> _______________________________________________
>> AMBER mailing list
>> AMBER.ambermd.org
>> http://lists.ambermd.org/mailman/listinfo/amber
>>
>
>
> --
> *Dr. Thomas Steinbrecher*
> Principal Scientist CADD
>
> Roche Pharma Research and Early Development
> Roche Innovation Center Basel
> F. Hoffmann-La Roche Ltd
> Bldg. 092/3.92
> Grenzacherstrasse 124
> 4070 Basel
> Switzerland
>
> Phone +41 61 682 1319
> mailto: thomas.steinbrecher.roche.com
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Thu Feb 16 2023 - 07:00:02 PST
