MacroModel Reference Manual - ISP

Appendix D: Force-Field File FormatAMBER n=2, γ=180: use V 2 /2 directlySome AMBER publications give torsional parameters as V n /2 and some as V n . In MacroModelforce-field files, always use the V n /2 value.We also implement AMBER torsional offsets explicitly for arbitrary offsets; however, theimplementation is not as efficient as that described in the preceding two paragraphs. Thus, forexample, you may, in principle, specify an n=2 potential using a 180˚ offset by means of apositive value of V n /2 and an explicit value of 180˚, but you are advised instead to use theformulation given above. On the other hand, for offsets equal to neither 0 nor 180˚, there isreally no choice but to use an explicit specification.To specify offsets explicitly for V 1 ,V 2 , and V 3 , use a continuation force-field line beginningwith 64 and place the offsets in the same columns as the V values for the main torsion line.Similarly, to specify offsets for V 4 ,V 5 , and V 6 , use a continuation line beginning with 74 andinsert the offsets in the same location. (Recall that V 4 ,V 5 , and V 6 themselves use a continuationline beginning with 54.) An offset specified as 0 is taken to mean “no offset specified,” sothat if you wish to use a V 2 potential with a zero offset and a V 3 potential with a 31˚ offset, usea negative value of V 2 /2, as described above, on the first torsion line for this interaction; placethe positive value of V 3 /2 on the same line. On a continuation line beginning with 64, indicatea V 2 offset of zero and a V 3 offset of 31˚.AMBER publications give total barriers for generalized torsional arrays—arrays with wildcard atoms (the X atom type in AMBER publications) at positions 1 and 4. MacroModel,however, uses the MM2 method of storing only the component torsional barriers except in thesubstructure part of the field. Thus only generalized AMBER torsional barriers should bedivided by the multiplicity of the torsional array where the multiplicity if equal to the numberof substituents on the 2nd atom (excluding the 3rd atom) multiplied times the number ofsubstituents on the 3rd atom (excluding the 2nd atom) in a torsional array of four atoms: 1-2-3-4. For the central bond of hexamethyl ethane for example, the multiplicity would be 3*3 or 9.For the peptide linkage, the multiplicity would be 2*2 or 4. Note that the multiplicity iscomputed based on the actual number of substituents so the multiplicity of a pair of unitedatom methylenes, e.g., x - CB - CB - x (MacroModel) or X - C2 - C2 - X (AMBER),would be only one since only single substituents (X) are attached to the methylene carbons. Ifyou are confused, compare some of the entries in the MacroModel AMBER force field withthose published in the AMBER papers. Remember that if a torsional array is described inAMBER publications without wild card X atoms, then the barriers are simply adjusted by signas described above, but not divided by multiplicity.Regarding AMBER-style generalized torsions, MacroModel follows the AMBER protocol ofnot including the multiplicity for generalized torsions in substructures. only. Thus the substructuretorsional barriers are the same as published in the AMBER papers if they are in the forceMacroModel 9.7 Reference Manual 209