3
$\begingroup$

In this paper (e.g. doi:10.1021/cs500093h and other applied QM/MM studies), it is demonstrated several transition states, intermediates, and products of given set of biomolecules (several residues, metal ions, and its substrates). What makes me it is interesting is that they depicted carefully the lose and gain of the atoms on each molecules. I have seen on some places that the showed bonds can be reproduced from some molecular docking softwares (e.g. Schrodinger Maestro, etc.) but it would be also interesting if a software could also draw the electron flow arrows (reaction mechanism) for each states. I have seen some research papers about docking studies before but they only depicted the bonds without giving any information of whether an atom had newly made or unreactive (the paper I cited clearly showed the lose of hydrogen and the bond forming on intermediate state).

How do they model the states? Is it that we have to manually draw the molecules on each states and put the protonated/deprotonated and dissociated/bonded forms combinatorically for the states and we evaluate the energy for each of these scenarios? Or the softwares could also make and break those bonds automatically and showed the energy for each of them? Which software? On the cited paper, I can not find any specific line that mentioning this and I am curious with the line on the paper, stated "the QM/MM boundaries crossed carbon-carbon bonds, and hydrogen atoms were used as “link” atoms". Is this one?

$\endgroup$

1 Answer 1

3
$\begingroup$

So let me summarize your questions: (1) is it possible for a software to automatically draw the electron flow arrows? (2) must the user add and remove the atoms manually, when modelling a TS?

The answer to (1): In principle yes, but usually a quantum chemistry program only outputs (a) the bond order between each pair of atoms, and (b) the charge of each atom. When you have these information for the reactant(s), the TS and the product(s), you know where the electrons are before, during, and after the reaction. You can then trivially draw the electron flow arrows. (With stepwise concerted mechanisms, you may need to calculate the bond orders and atomic charges of additional structures along the IRC in order to correctly draw the arrows, but this is a minor technicality.) Nevertheless, I'm not aware of any software that can do the last arrow-drawing step automatically, partly because it is easy enough for humans to do.

The answer to (2): Sometimes, but not always. The most straightforward, although not very robust, approach is to come up with a guess of the TS structure and then optimize it with a quantum chemistry program. This is prone to non-convergence unless you have at least some chemical intuition about how a TS should look like. A second class of methods (e.g. QST2, NEB) is to automatically find the TS through two user-supplied structures, namely the reactant complex and the product complex; these can be made more robust if the user can also specify a rough guess of the TS structure (as in the QST3 method). Still another class of methods (e.g. GRRM) only require that you supply a reactant structure, and will automatically find all possible reactions from this reactant, as well as the associated transition states; however this is extremely costly, since the program has no idea about which reactions you are mostly interested in, so most of the time are wasted in finding totally uninteresting TSs, like the TSs of methyl rotations.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .