I ran MD of a Prot-Lig complex for 200ns.

Then proceeded to GBSA calculation using gmx_MMPBSA.

Start frame = 0 and Endframe = 20000; interval = 50 (in mmpba.in)

The entropy curve for the full time range showed subsequent changes, and

interaction entropy at 10,000th frame is around 11 kcal/mol. (See attached image)

However, when I intended to generate interaction entropy for 10,000th to 20,000th frame, using the subset of the data, the new curve, starts from zero kcal/mol.

Isn't it supposed to start from 11 kcal/mol?