Effect of correcting a fault in MOZYME

Fault

In a small number of systems, one or more non-bonded interactions were not included in the MOZYME localized molecular orbitals.  This resulted in calculated heats of formation that were lower than the equivalent heats of formation calculated using conventional MOPAC.  Jobs run using MOZYME and conventional methods should produce the same results.  If they are significantly different, then there is a fault in the MOZYME method..

Implications

In projects where heats of formation are used, typically for finding changes in energy of various related systems, the corrected MOPAC might produce heats of formation that would be different to those from older copies of MOPAC.  If that occurs, all energy changes calculated from using results from both the corrected and older MOPAC's would be meaningless.

Test for the fault

Only a small number of systems are likely to be affected. Testing to determine of a system is affected is straightforward.  Using a corrected version of MOPAC, one made after 18 August 2017, edit an ARC file for a system of interest to make a test data-set, and run two 1SCF calculations using that data-set, one using the old MOPAC and one using the new MOPAC.  If the resulting heats of formation are within 0.1 kcal/mole of each other, then the project would not have been affected by the fault.  If the difference is greater, take an appropriate corrective action. 

Corrective action

For projects that use heats of formation, started before 19 August 2017, do NOT upgrade to the Version 17.231 or higher until the project is complete.  If necessary, request a new license key, and include a note saying  that you want to continue using an old copy of MOPAC. If possible, include a comment on when the project would be completed.

Alternatively, use the ARC files and re-run all systems for which heats of formation are important, using a copy of MOPAC that was made after 19 August 2017, i.e., with a version number of 17.231 or higher, and include keyword 1SCF.