Discuss Jacobian term in the doc

[giacomofiorin]

See #367 #407 and some messages in the NAMD and LAMMPS forums.

Nothing too verbose or extensive is needed: users just ought to be aware that the term is present in a PMF computed with any method, but the term is known for some variables based on their total-forces implementation. Recent reviews, including those about ABF, tend to focus on newer methodology, making this piece of information harder to find.

Aside from notes exchanged in those issues/PRs, I would only add here that the appendix in the Mol Phys paper lists the Jacobian term in the total force, not in the PMF. Of course, the term can be integrated but a user needs to know how to do it. We could probably just have an expression for a post-hoc correction to the PMF.

[jhenin]

The places I can see in the doc where that could fit is either in the ABF section where the Jacobian term is already mentioned, or in the Thermodynamic integration section but it would have to be added there.

[giacomofiorin]

I'm thinking about a third option, i.e. a very brief intro section at the beginning of the chapter on biases, mentioning TI and other estimators (and the methods that use them). The Jacobian term is present on all these methods, including WHAM, M-BAR, metadynamics, ... So if you only mention it for some of the methods, you're implying that it does not apply to others when it actually does.

[jhenin]

Looked into it, and I found it difficult to write. Can you take a stab at it, if you have a precise idea what to say?