[gmx-users] dummy atom fep topology

[chris liopiri]

Dear All,

I wanna perform TI and use dummy atoms. I have seen that a lot of people use in the dummy topology bonded parameters like in the original system. Lets say we have a c3-hc that goes to c3-du then in dummy topology the c3-du bond will be the same as in c3-hc. But in this way my question is that going from dummy to the original system in LAMDA 0.0 the c3 will feel the mass of the du since the bond potential term will not be zero but equal to c3-hc. So is it more correct to set this potential term to zero and is this problem for sampling in case I have more atoms that I set as dummies?

looking forward for a reply,
chris