Questions about the inconsistent results of TI, BAR and MBAR by using alchemical-analysis

[lpfSTX]

Dear all,

The current alchemical-analysis software gave me inconsistent free energy values from TI, BAR and MBAR methods when I performed absolute hydration calculations for some small molecules. As I know, the alchemical-analysis software imports the pymbar software. In the pymbar, it uses the kT as its energy unit. In the alchemical-analysis, it defines kB (Boltzmann’s constant kJ/mol/K) and converts kJ/mol to kT. However, AMBER uses kcal/mol as its energy unit. The current alchemical-analysis version did not convert kcal/mol to kJ/mol for using AMBER output as input, which results in this inconsistency problem.

So I scale up the energy terms (dhdlt and u_klt) with multiplying 4.184 at the bottom of parser_amber.py and change the line 130 "beta_report = 4.184**(not b_kcal)*beta" to "beta_report = 4.184**(b_kcal)*beta" in alchemical-analysis.py. Then it can give me consistent free energy values from TI, BAR and MBAR for AMBER output analysis.

Best,
Pengfei