I plan to use CMAP terms in a MD simulation, I know that both top_all22_prot_cmap.inp and par_all22_prot_cmap.inp should be used, but besides them, is it necessary to explicitly specify "CMAP" somewhere in the input charmm script? Thanks in advance.
Reading the RTF/PARAM files with the CMAP definitions is sufficient, provided the CHARMM release is recent enough. I believe CMAP is included as a standard feature for c31b1 and later, and may be available as an optional feature (recompile required) for c30 versions.
Thank you very much for your fast reply. I recently perofrmed some MD simulations for ace-ala-nme in tip3 water, I did read the RTF/PARAM files with the CMAP definitions, but for the all 5ns simulation started from different initial conformation of ala2, the results of fi-psi angle population are consistent to the PAR22 result, but not the CMAP result. That is why I would like to ask the above question.
Here are one of my results, starting from same initial conformation of ALA2, but at different temperature, the left panel corresponds to T=300K, and the right panel corresponds to T=600K. They are typical PAR22 rather than CMAP field results.
The above doesn't apply at chain ends, however, as the atom names for the acetyl carbon and methylamine N are different. There is a custom residue for the alanine dipeptide, ALAD, which has the CMAP terms applied properly for this somewhat pathological case. There are comments in the topology file that indicate this.