Cryo-EM, as it is currently practiced in many laboratories, is limited to the visualization of molecules that are in thermal equilibrium at the time before freezing. A further limitation is that the existing software does not fully exploit the information that is contained in the images of large ensembles of molecules in thermal equilibrium. This book is a collection of recent articles by the author, reprinted with introductions, and they mainly describe two novel methods in cryo-EM, one computational and the other experimental that requires the use of a microfluidic device. Both methods have the capacity to shed light on the dynamic behavior of biomolecules. Combined, they greatly expand the range of applications of cryo-EM.

The book describes a successful approach in which, based on cryo-EM data, all states visited by the molecule in thermal equilibrium are mapped by manifold embedding--a method of geometric machine learning--and the energy landscape of the molecule is derived. It also discusses methods and biological results of time-resolved cryo-EM, following a reaction in a non-equilibrium system over a short period of time and resulting in the capture of short-lived states that have been inaccessible by standard methods of cryo-EM.