TURBOMOLE Users Forum

Forum General => TURBOMOLE Forum General => Topic started by: jbaltrus on June 26, 2011, 01:48:41 AM

Title: QM/MM, e.g. non periodic charge embedding?
Post by: jbaltrus on June 26, 2011, 01:48:41 AM
Is there a way to perform non-periodic point charge embedding, e.g. some sort of QM/MM using PEECM model? Seems like evgenyi below has similar question, e.g. there is always this problem when the system becomes too large even for TURBOMOLE and some of it needs to be treated at a very low level of theory (e.g. QM/MM), I was wondering if PEECM can be used for that?

thanks
Title: Re: QM/MM, e.g. non periodic charge embedding?
Post by: jbaltrus on June 26, 2011, 02:03:50 AM
furthermore, I found the paper that talks about interface between TURBOMOLE and CHARM for these type of calcs. Has that ever been implemented?

thanks

The implementation of a hybrid QM–MM approach combining ab initio and density functional methods of TURBOMOLE with the molecular mechanics program package CHARMM is described. An interface has been created to allow data exchange between the two applications. With this method the efficient multiprocessor capabilities of TURBOMOLE can be utilized with CHARMM running as a single processor application. Therefore, features of nonparallel running code in CHARMM like the TRAVEL module for locating saddle points or VIBRAN for the calculation of second derivatives can be exploited by running the CPU intensive QM calculations in parallel. To test the methodology, several small systems are studied with both Hartree–Fock and density functional methods and varying QM–MM boundaries. Also, the computationally efficient RI-J method has been examined for use in QM–MM applications. A B12 cofactor containing cobalt has been studied, to examine systems with a large QM region and transition metals. All tested methods perform satisfactory in comparison with pure quantum calculations. Additionally, algorithms for the characterization of saddle points have been tested for their potential use in QM–MM problems. The TRAVEL module of CHARMM has been applied to the Menshutkin reaction in the condensed phase, and a saddle point was located. This saddle point was verified by calculation of a steepest descent path connecting educt, transition state, and product, and by calculation of vibrational modes. © 2003 Wiley Periodicals, Inc. J Comput Chem 24: 1240–1249, 2003
Title: Re: QM/MM, e.g. non periodic charge embedding?
Post by: mas on June 28, 2011, 07:38:53 PM
The PECCM method is by definition working within the periodic boundary conditions. For the non-periodic case it simply reduces to usual point charges embedding. There is a way to efficiently do QM/MM calculations using milions and even more of point charges, but it is not yet implemented in Turbomole. The implementation would be pretty straightforward but I was not sure if there was a market for such calculations. Does anybody has a problem where the point charges embedding is the most time consuming step?
Title: Re: QM/MM, e.g. non periodic charge embedding?
Post by: jbaltrus on June 28, 2011, 08:02:02 PM
My two cents: TURBOMOLE is inherently non periodic software so from the first look non periodic charge embedding would be of more use (ONIOM-like???)
Title: Re: QM/MM, e.g. non periodic charge embedding?
Post by: mas on June 28, 2011, 09:35:03 PM
The non-periodic embedding using point charges has been available in Turbomole for years. The ONIOM method, at least the one implemented in (you know where) by default does not use point charges embedding but just a subtraction scheme. You could probably force the ONIOM method to use point charges but I haven't seen any reasonable application for this yet.
BTW, we have a periodic extension of Turbomole, at least for the DFT GGA and LDA energies, which will be available in the future.
Title: Re: QM/MM, e.g. non periodic charge embedding?
Post by: uwe on June 28, 2011, 09:39:37 PM
Hi,

if you have really really many point charges, 99% of the time the Turbomole routines are busy with symmetry checks and with the automatic option to check for too close lying point charges. Also the self interaction energy is always calculated but only printed if requested.

The next Turbomole release will have an option to avoid all that and just take care of the terms needed for usual QM/MM (no symmetry, point charges come from FF, so there will be no two such points at the same position, no self-energy is needed, etc.). That will make the treatment of million(s) point charges much cheaper.

Regards,

Uwe
Title: Re: QM/MM, e.g. non periodic charge embedding?
Post by: jbaltrus on June 28, 2011, 09:57:59 PM
looking forward to it