From kB to kB: Entropy evaluation made easy

TYPEStatistical & Bio Seminar
Speaker:Prof. Roy Beck-Barkai
Affiliation:Tel Aviv University
Organizer:Yariv Kafri
Date:14.04.2019
Time:14:30 - 15:30
Location:Lidow Nathan Rosen (300)
Abstract:
Entropy and free-energy estimation are key in thermodynamic characterization of simulated systems ranging from spin models through polymers, colloids, protein structure, and drug-design. Current techniques suffer from being model specific, requiring abundant computation resources and simulation at conditions far from the studied realization. In this talk, I will present a novel universal scheme to calculate entropy using lossless compression algorithms and validate it on simulated systems of increasing complexity [1]. Our results show accurate entropy values compared to benchmark calculations while being computationally effective. In molecular-dynamics simulations of protein folding, we exhibit unmatched detection capability of the folded states by measuring previously undetectable entropy fluctuations along the simulation timeline. Such entropy evaluation opens a new window onto the dynamics of complex systems and allows efficient free-energy calculations.

[1] R. Avinery, M. Kornreich, R. Beck, Universal and efficient entropy estimation using a compression algorithm, under review (2018). ArXiv:1709.10164
 
 
Prof. Roy Beck-Barkai