How to decide which force field will be appropriate for a certain protein in a Gromacs simulation?
There is a lot of literature about this, comparing different force-fields and trajectory times. As a first approach I would suggest you to choose the most used forcefields like 43a1 (Gromacs) or 99SB (Amber). In a very general point of view all programs compute the same equations of motion and the ff take into account the same interactions...I my opinion the ff is the the weakest link in the chain! For this I recommend to use the most tested force fields. This are a nice papers about:
http://www.sciencedirect.com/science/article/pii/S0006349508702777
http://www.sciencedirect.com/science/article/pii/S0006349510005606
There is a lot of literature about this, comparing different force-fields and trajectory times. As a first approach I would suggest you to choose the most used forcefields like 43a1 (Gromacs) or 99SB (Amber). In a very general point of view all programs compute the same equations of motion and the ff take into account the same interactions...I my opinion the ff is the the weakest link in the chain! For this I recommend to use the most tested force fields. This are a nice papers about:
http://www.sciencedirect.com/science/article/pii/S0006349508702777
http://www.sciencedirect.com/science/article/pii/S0006349510005606
Diego makes good points. I will just expand on his response to say you need to make sure of the context. Nearly all MM force fields are reasonably accurate for a folded protein, though subtle effects may differ. If you're studying protein folding, disordered states, etc. there are definitely some force fields that are better than others. There is a ton of literature on this topic.
Choosing ff is tricky. You have to read literatures. It's the only way.
Selecting FF is very important and it depends on which question you would like to answer. It is always better to read manuscripts based on your context to know about the FF. If you couldn't figure out, then in my opinion, it is better to use GROMOS96 54a7, 43a2 or AMBER99SB-ILDN. Like Dr. Lemkul said all FF are accurate for folded proteins. Having said that, if you are planning to simulate any Post-Translationally Modified (PTM) proteins, then make sure the modified residues are in .rtp registry.
- Badges
- Science topic