Abstract.

We measure the free energy of a model filament, which undergoes deformations and structural transitions, as a function of its extension, in silico. We perform Brownian Dynamics (BD) simulations of pulling experiments at various speeds, following a protocol close to experimental ones. The results from the fluctuation theorems are compared with the estimates from Monte Carlo (MC) simulation, where the rugged free energy landscape is produced by the density of states method. The fluctuation theorems (FT) give accurate estimates of the free energy up to moderate pulling speeds. At higher pulling speeds, the work distributions do not efficiently sample the domain of small work and FT slightly overestimates free energy. In order to comprehend the differences, we analyze the work distributions from the BD simulations in the framework of trajectory thermodynamics and propose the generalized fluctuation theorems that take into account the information (relative entropy) evaluated in the expanded phase space. The measured work - free energy relation is consistent with the results obtained from the generalized fluctuation theorems. We discuss operational methods to improve the estimates at high pulling speed.

Graphical abstract

中文翻译：

---

用广义涨落定理测量自由能：模型灯丝的理论和数值研究

摘要。

我们测量模型灯丝，其经历变形和结构转变的自由能，作为其扩展的功能，在计算机芯片上。我们遵循接近实验的协议，以不同的速度执行拉动实验的布朗动力学（BD）仿真。将波动定理的结果与蒙特卡洛（MC）模拟的估计值进行比较，在蒙特卡洛（MC）模拟中，通过状态密度方法产生了崎free的自由能态。波动定理（FT）给出了在中等牵引速度下自由能的准确估计。在较高的牵引速度下，功分布不能有效地采样小功的范围，而FT会稍微高估自由能。为了理解差异，我们在轨迹热力学的框架内分析了BD模拟的功分布，并提出了考虑了在扩展相空间中评估的信息（相对熵）的广义涨落定理。测得的功-自由能关系与从广义涨落定理获得的结果一致。我们讨论了提高牵引速度时估算值的操作方法。

图形概要