Hi Wojciech,

Thanks a lot for your answer! Please, note that I was referring to a case where one is performing an alchemical transformation of one ligand into another one, in which there are some atoms whose charges are turned off and there are other new atoms that are grown and their charges are gradually turned on. This is a context in which one seeks to calculate relative binding free energies rather than absolute ones.

I understand from your explanation that if one scales the charge of a neutral molecule with lambda, it will remain neutral all along, as lambda will be a prefactor multiplying a summation of charges that is zero. However, I can't see right now how that is true also for the case when one is turning off the charges of certain atoms while leaving the remaining ones unchanged (decharging process), as required in relative free energy calculations. The same is true for the independent recharging process, when one charges the newly created atoms.

I have been thinking if the actual answer is that the bias introduced in the decharging process because of the variation of net charge is cancelled by the one originated in the recharging step, as the charges of the initial and final states will effectively be the same if no net charges are created. If that is the case, then I find the statement at alchemistry.org for the transformation of toluene into phenol (step 1 in lambda intermediates section) a bit misleading, as the net charge will not be conserved for all lambda intermediates during the decharging process.

Cheers,

Jorge

Jorge,

I can see how the toluene/phenol example might come off as misleading. However, the author of the tutorial clearly states that the total charge should be kept constant throughout intermediate steps; if you want to adhere to the outlined procedure, this can easily be done by e.g. uniformly distributing the excess charge over the remaining non-dummy atoms of the ligand in steps 2 and 3, as the intermediates need not correspond to any chemically meaningful entity.

Best,

Miłosz

Hi Milosz,

What you say is indeed a nice solution. However, in this tutorial http://ambermd.org/tutorials/advanced/tutorial9/ that calculates the relative free energy of binding of toluene and phenol to T4 lysozyme mutant L99A, the excess charge is not redistributed, as can be easily checked by taking a look to the .lib files of the compounds. I would like to know, then , whether the principle of constant charge throughout all lambda intermediates is a sine qua non condition for accurate relative free energy calculations.

Kind regards,

Jorge