Enhanced sampling techniques have revolutionized molecular
dynamics
(MD) simulations, enabling the study of rare events and the calculation
of free energy differences in complex systems. One of the main families
of enhanced sampling techniques uses physical degrees of freedom called
collective variables (CVs) to accelerate a system’s dynamics
and recover the original system’s statistics. However, encoding
all the relevant degrees of freedom in a limited number of CVs is
challenging, particularly in large biophysical systems. Another category
of techniques, such as parallel tempering, simulates multiple replicas
of the system in parallel, without requiring CVs. However, these methods
may explore less relevant high-energy portions of the phase space
and become computationally expensive for large systems. To overcome
the limitations of both approaches, we propose a replica exchange
method called OneOPES that combines the power of multireplica simulations
and CV-based enhanced sampling. This method efficiently accelerates
the phase space sampling without the need for ideal CVs, extensive
parameters fine tuning nor the use of a large number of replicas,
as demonstrated by its successful applications to protein–ligand
binding and protein folding benchmark systems. Our approach shows
promise as a new direction in the development of enhanced sampling
techniques for molecular dynamics simulations, providing an efficient
and robust framework for the study of complex and unexplored problems.