Can you explain these: - normal mode basis? - TLS – translation libration screw??

let us now consider the important point – how can we reconcile the infitesimally small variations across the conformational states up and down the hierarchy with the materialization at each level. That is, i quite like all the ideas we propose here and they capture quite a lot, but what is still bothering me is the case where we have some local heterogeneity at the few residues level (b-factors) and then also some heterogeneity at the higher-component level (like the chain it belongs to). How do we jointly represent this via our containmnent and heterogeneity regimes? What if the residues/atoms are like 0.05 Angstrom apart in each states, but there are a 100 states and each atom/residue is going their own way whereas the chain component is varying across just a few axes (TLS)? Do we just programmatically delta/diff encode this at one level of the hierachy, but then assign states to the other? Can we mix and match heterogeneity regimes i guess is what i’m asking? How does the DAG idea of states fit into this?

the strenght of annotations is in the fact that they can refer to precompiled molecular queries/selectors so one can tag arbitrary groupings/arrangements of entities(atoms, chains, domains, subsets, pairs thereof etc.) either in cartesian space or across state space or both.

Presentation order: