In case of studying protein-ligand interaction theoretically people generally use three methods to determine binding free energy, a) MMPBSA-GBSA method, b) linear interaction energy (LIE) and c) thermodynamic integration (TI).
It would be great if people could comment on the possible merit and demerits of those methods and provide protocols to follow.
First, this classification is very arbitrary and incomplete. Second, it depends on your definition of better. Accurate: smallest max error or mean error? Computational cost? The most physically rigorous?
Also, I commented already on this, but the key in the binding free energy calculation is the underlying thermodynamic cycle. Then you just introduce approximations in evaluating the relevant free energy contributions, which are mostly dependent on how much computer time you can afford to spend. As usually, everything comes at a price.
Btw, TI is just the free energy estimator for the free energy perturbation and in fact LIE is based on the end-point TI (aka LRA).
For protocols on a) Peter Kollman' s articles describe the methodology b) is Johan Aqvist and c) and in general works by Arieh Warshel. Also more recent papers on SAMPL challenge are pretty interesting.
https://www.researchgate.net/post/Anyone_familiar_with_binding_free_energy_of_an_ion_from_alchemical_transformation
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry
- Computational Chemistry
- Molecular Dynamics