So Shunzhou suggested that it could be a problem:
"I think the last definition will be used in these cases where duplicate lines exist.
In cases where multiple terms are used to define a same dihedral (multiple terms with the last one having a negative period), it may affect the energies when replicates appear."
I checked all the dataset (5 proteins) and I have not found a single case with a negative periodicity in the dihedral.
According to the documentation:
http://ambermd.org/FileFormats.php#frcmod
PN The periodicity of the torsional barrier.
NOTE: If PN .lt. 0.0 then the torsional potential
is assumed to have more than one term, and the
values of the rest of the terms are read from the
next cards until a positive PN is encountered. The
negative value of pn is used only for identifying
the existence of the next term and only the
absolute value of PN is kept.
Based on this description I am not sure what to conclude.
If we ever run into the negative case, we'll test it.
So Shunzhou suggested that it could be a problem: "I think the last definition will be used in these cases where duplicate lines exist.
In cases where multiple terms are used to define a same dihedral (multiple terms with the last one having a negative period), it may affect the energies when replicates appear."
I checked all the dataset (5 proteins) and I have not found a single case with a negative periodicity in the dihedral.
According to the documentation: http://ambermd.org/FileFormats.php#frcmod PN The periodicity of the torsional barrier. NOTE: If PN .lt. 0.0 then the torsional potential is assumed to have more than one term, and the values of the rest of the terms are read from the next cards until a positive PN is encountered. The negative value of pn is used only for identifying the existence of the next term and only the absolute value of PN is kept.
Based on this description I am not sure what to conclude.
If we ever run into the negative case, we'll test it.