Dears,

I have a couple of questions regarding the application of Metadynamics to study protein dimerization. I am not a physicist and I have no previous experience with this technic so my questions might be very basic and I would really appreciate your help and clarifications.

Let’s say I have 2 different conformations of a dimer (same protein): one is what we think is the good dimer (good association on the monomers) and the second one is “wrong”. We want to use metadynamics to see how the dimerization occurs and to get a view of the energy landscape of the protein that can help us (hopefully) at the end to say: this is a good/stable model and this is not.

1- My first question is: Is metadynamics the best method that can help me to answer this question. Is there any variants (of metadynamics) that are more suitable to my case than others?

2- Metadynamics implies choosing Collective Variables (CVs). This latter should respect some guidelines (from "Laio, A. et al. Metadynamics: a method to simulate rare events and reconstruct the free energy in biophysics, chemistry and material science.):

- “They should distinguish between the initial and final state and describe all the relevant intermediates” —> What if I don’t have enough information about my process? I only know about my initial and final states (monomer to dimer or vice versa) but no information about the intermediates states! I would love to have explanation/examples about “intermediate state”, maybe I misunderstood it.

- “They should include all the slow modes of the system” —> This point is also not very clear to me. What are the slow modes and how can I identify them?

3- I have unbiased Coarse-Grained trajectories (Martini FF) of the same protein in a dimeric and monomeric apo-form (good/bad monomer/dimer). Is it possible to apply dimensionality reduction methods on these trajectories in order to retrieve information about the CVs that I could use?

NB: I couldn’t see any transitions from monomer to dimer (or vice versa) from these trajectories… which is expected since they are unbiased and not too long (~4 microseconds).

4- From experimental data, I have information about the residues that should be at the interface of the dimer. I was thinking to use the distances between the COM of these residues in each monomer as CVs. Is it a reasonable choice? for me it is but as I said, I have no previous experience with metadynamics so I am constantly questioning my protocol.

Thank you in advance for your help and clarifications.