Jumpcount: Estimation of Uncertainties in Biomolecular Simulations

Pavel Kříž1, Jan Beránek2, Vojtěch Spiwok2

1Faculty of Mathematics and Physics, Charles University, Prague, Czech Republic
2Department of Biochemistry and Microbiology, University of Chemistry and Technology, Prague, Czech Republic

Increasing number of biomolecular simulations provide sufficiently long simulations to predict equilibrium constants and associated free energies of simulated processes such as protein folding or protein-ligand binding. The equilibrium constant of a process of a transition from the state A to state B can be estimated as the fraction of times spent in B and A. Here we present a simple, yet exact, method to calculate the errors of the free energy estimations, which is based solely on temperature and the number of transitions from A to B and B to A. The 95% confidence interval of free energy calculated from a trajectory with one A to B and one B to A is the value ± 9.1 kJ/mol (2.2 kcal/mol) at 300 K. For two A to B and B to A transitions it is the value ± 5.7 kJ/mol (1.4 kcal/mol). Markovianity of the process is the prerequisite of the method. The error can be calculated online by our tool at jumpcount.cz.

The work was supported by Czech Science Foundation (22-29667S and 19-16857S) and the Ministry of Education, Youth and Sports of the Czech Republic (LM2018140, LM2018131).