Home   Manual

Metals in MMFF94

Good Morning!
I would like to know, currently which metals are parameterized in the MMFF94 and MMFF94s and if there are probable transition metals in a parameterization study?

The parameters for MMFF94 are in the papers, e.g.
Halgren, T.A. (1996), Merck molecular force field. I. Basis, form, scope, parameterization, and performance of MMFF94. J. Comput. Chem., 17: 490-519

I don’t remember off the top of my head, but I think it’s limited only to a few ions:

   FE+2   87   87   87   87   87    IRON +2 CATION
   FE+3   88   88   88   88   88    IRON +3 CATION
   LI+    92   92   92   92   92    LITHIUM CATION         
   NA+    93   93   93   93   93    SODIUM CATION          
   K+     94   94   94   94   94    POTASSIUM CATION       
   ZN+2   95   95   95   95   95    DIPOSITIVE ZINC CATION 
   CA+2   96   96   96   96   96    DIPOSITIVE CALCIUM CATION
   CU+1   97   97   97   97   97    MONOPOSITIVE COPPER CATION
   CU+2   98   98   98   98   98    DIPOSITIVE COPPER CATION
   MG+2   99   99   99   99   99    DIPOSITIVE MAGNESIUM CATION

In general, if you want metal complexes, I’d recommend UFF as a force field.

Whenever I tried including any of these ions in the system, the optimization in Avogadro with MMFF94 or MMFF94s returned the error: “AutoOpt: Could not setup force field …”. Other FFs (UFF, GAFF, Ghemical) are fine. Any idea? Does that mean Avogadro doesn’t have the MMFF94 parameters for any metal?

It has all the MMFF94 parameters, but is likely not assigning the atom types.

If GAFF is working, I’d go with that - David van der Spoel’s group has spent time checking the parameter assignments. UFF is also a possibility.

I understand that in case the metal ion have different charge (Fe+2, Fe+3, Cu+1, Cu+2) the atom types might be necessary and it could be a problem. But for non transition metal there is no ambiguity. If it is an easy job maybe Avogadro should at least get them correctly.

I’d need to see your system.

The code uses Open Babel to do atom typing and force field optimization, e.g. atom typing for potassium

Certainly the simple cations (Li, Na, K, Zn, Ca, Mg) work, e.g. solvating a \ce{Ca+2} ion with \ce{H2O}