I received your files. The thing is the variable 'assign_chainID' is ignored when you pass a reference structure... in this case, your reference structure doesn't contain the chain IDs either and that's why gmx_MMPBSA end up in error... just add the chain IDs to your reference structure 'md.pdb' (A for the protein and B for the ligand for instance) and also make sure your
mmpba.in file has the proper notation for the mutant (see bold letters). I also see you are specifying 'startframe=9900, endframe=10000', remember this refers to frames and not time, so if your trajectory has 10000 frames that's ok... if not make sure 'startframe' and 'endframe' are defined according to the number of frames you are analyzing in the trajectory...
Sample input file for entropy calculations
#This input file is meant to show only that gmx_MMPBSA works. Althought, we tried to used the input files as recommended in the
#Amber manual, some parameters have been changed to perform more expensive calculations. Feel free to change the parameters
#according to what is better for your system.
&general
#
startframe=9900, endframe=10000, interval=1,assign_chainID = 1
protein_forcefield="oldff/leaprc.ff99SB",
/
&gb
igb=2, saltcon=0.150,intdiel=8
/
&alanine_scanning
mutant='ALA'
mutant_res='A:240'
/
hope this helps!
cheers!