Time matters: timescales determine the fate, behavior and functionality of living matter. Especially the interplay between fast and slow molecular processes is omnipresent in biochemistry (protein folding, enzymatic conversions, molecular signaling, etc.). However, in chemistry, models are lacking to properly understand time and its effect on this difficult biological reality.

PASTIME will lay the methodological foundations for a correct understanding of time in chemistry by introducing memory dependent transition interface path sampling to chemical modeling. We will tune the memory that is kept in simulations by creating new path ensemble definitions. The problem will be made computationally solvable by cutting molecules’ pathways short. Long timescale effects will be studied through pathway statistics. We will achieve higher decorrelation in the sampling by designing new Monte Carlo moves in path space, such as alchemical moves that change the identity of atoms on the fly, and smart use of memory expansion with a replica exchange move. These new time concepts will be challenged and experimentally validated in a context of drug transport kinetics: drug (un)binding to protein binding sites, and permeation of molecules through cell membranes. The modeling principles will be converted to new algorithms that are computationally feasible today and will be shared in open software.

With PASTIME we will be able to understand slow time effects in molecular interactions. Appropriate memory reduction and tuning will give us the methods to understand time in an endless range of molecular processes, and this work will contribute to drug design, biochemistry, physiology, catalysis, material and polymer science.