alchemlyb: NAMD parser does not take temperature as argument
The NAMD parser #7 (for FEP) does not take T as an argument and does not return u_nk in proper units of kT.
About this issue
- Original URL
- State: closed
- Created 5 years ago
- Comments: 19 (9 by maintainers)
Commits related to this issue
- add units of energy and time to the descriptions of the datasets - Gromacs - Amber - NAMD - see also https://github.com/alchemistry/alchemlyb/issues/75 — committed to alchemistry/alchemtest by orbeckst 5 years ago
- WIP: trying to make NAMD parser return energies in kT (see #75) — committed to alchemistry/alchemlyb by orbeckst 5 years ago
- add units of energy and time to the descriptions of the datasets - Gromacs - Amber - NAMD - see also https://github.com/alchemistry/alchemlyb/issues/75 — committed to alchemistry/alchemtest by orbeckst 5 years ago
- new reference value for NAMD BAR "tut03_tyr2ala (hydrated)" dataset - changed the test reference value to the value produced by the NAMD parser + BAR (alchemlyb) implementation - see https://github... — committed to alchemistry/alchemlyb by orbeckst 5 years ago
- WIP: trying to make NAMD parser return energies in kT (see #75) — committed to alchemistry/alchemlyb by orbeckst 5 years ago
- new reference value for NAMD BAR "tut03_tyr2ala (hydrated)" dataset - changed the test reference value to the value produced by the NAMD parser + BAR (alchemlyb) implementation - see https://github... — committed to alchemistry/alchemlyb by orbeckst 5 years ago
- WIP: trying to make NAMD parser return energies in kT (see #75) — committed to alchemistry/alchemlyb by orbeckst 5 years ago
- new reference value for NAMD BAR "tut03_tyr2ala (hydrated)" dataset - changed the test reference value to the value produced by the NAMD parser + BAR (alchemlyb) implementation - see https://github... — committed to alchemistry/alchemlyb by orbeckst 5 years ago
The BAR and MBAR operate on dimensionless quantities: kT must be in the same units as the energies. Otherwise beta*U becomes something different. . . If BOTH the energies and kT change units, then everything is valid, except of course the free energies will be in those different units.
Currently the software modules assume you are passing in the beta*U quantities (already nondimensionalized). So you multiply the results by whatever kT you divided the energies by before you passed them into BAR.
Thanks for confirming.
Currently, the temperature from the output file is not used, is it?
I don’t think that it would be correct to use the instantaneous temperature for kT – the Boltzmann factor comes from a Lagrange multiplier for system at constant T (i.e., constant average temperature) so this T should be the “target temperature” of the thermostat in the simulation. But maybe @mrshirts @davidlmobley want to correct me here.