Research

Recently, we have addressed the relevant and timely issue of correctly evaluating energy dissipation of non-equilibrium stochastic systems with timescale separation, when having access to only a subset of variables. Despite this information loss, we developed a scheme to reveal such hidden dissipation by studying the violation of the Fluctuation-Dissipation Theorem for an observed slow variable. In the frequency domain, we prove generally that the violation spectrum exhibits a vanishingly small plateau in the intermediate frequency region due to the non-equilibrium coupling of the observed slow variable with the hidden fast one. Surprisingly, the integral of this small plateau preserves the finite dissipative information even in the timescale separation limit. We show full retrieval of hidden dissipation in an analytical model that describes a particle subject to a fast and random switching of the energy landscape. We believe that our work is very useful experimentally to identify the corresponding hidden active processes. The work is published in Phys. Rev. Lett. in 2016 [1].

[1] Shou-Wen Wang, Kyogo Kawaguchi, Shin-ichi Sasa, and Lei-Han Tang. Entropy production of nanosystems with time scale separation. Phys. Rev. Lett., 117:070601, Aug 2016.

I have a following up preprint that discusses the how to infer dissipation from violation of the fluctuation-dissipation theorem when the Markov process has an asymmetric load-sharing factor [2].

[2] Shou-Wen Wang. Inferring dissipation from violation of the Fluctuation-Dissipation Theorem, Phys. Rev. E 97, 052125 (2018)

We also show surprisingly that energy can be extracted from systems via a periodic external field, when such a system could adapt to the slow variation of the external field. This is a proto-type of frequency engines.

[3] Shou-Wen Wang, and L.-H. Tang, Extracting energy from non-equilibrium fluctuations without using information, arXiv preprint arXiv:1611.04089

I am also interested in how biological processes exploit energy dissipation to detect the environment more accurately, which we analyzed recently for a relevant model of sensory adaptation in E.coli [4]. The ability to monitor nutrient and other environmental conditions with high sensitivity is crucial for cell growth and survival. Sensory adaptation allows a cell to recover its sensitivity after a transient response to a shift in the strength of extracellular stimulus. Recently, G. Lan et al. performed a detailed analysis of a stochastic model for the E. coli sensory network. They showed that accurate adaptation is possible only by dissipating chemical energy even at the stationary state. We are therefore motivated to understand how signal transduction motif can use chemical energy to do a better job. We have performed exact analysis on the model proposed by Lan et al. Our study suggests that the adaptation error can be reduced exponentially as the methylation range increases. Finally, we show that a nonequilibrium phase transition exists in the infinite methylation range limit, despite the fact that the model contains only two discrete variables.

[4] Shou-Wen Wang, Yueheng Lan, and Lei-Han Tang. Energy dissipation in an adaptive molecular circuit. J. Stat. Mech., 2015(7):P07025, 2015.

Pulsation plays an essential role in coordinating multicellular developmental programs but also appears in other biological contexts. To enter into collective oscillations, cells communicate with each other through chemical or mechanical cues [See Fig.~3(a)]. The underlying biochemical network to process/dispatch extracellular signal is under intense experimental investigation. In this work, we shift our attention from details of the molecular construct to the whole-cell response characteristics and energy flow. Using response theory and thermodynamic arguments, we show that, under periodic stimulation, cells must develop a phase-lead (a form of anticipatory behavior) to enable energy outflow to drive oscillations in a passive medium. Using Kramers-Krönig relation (an intrinsice relation for response function simply due to causality), we further demonstrate that this necessary condition for autonomous behavior is met by cells whose signal dispatch rate adapts to a shift in the incoming signal strength [See Fig.~3(b) and (c)]. Revisiting various detailed models and experimental systems, we find adaptation to be a unifying feature behind “dynamical quorum sensing” reported in the recent literature. Our study led to a set of new hypotheses whose experimental validation should expedite the quest for the origin of pulsation in natural biological systems, and aid the design of clock in synthetic systems. Our study also demonstrated that auto-induced oscillation in cell populations is a particularly simple phenomenon that affords a low-dimensional description, and it establishes a new paradigm for collective oscillations that is different from the well-known Kuramoto scenario where individual cells need to oscillate spontaneously.

[5] Shou-Wen Wang and Lei-Han Tang, Adaptation route to emergent oscillations in communicating cell populations, (in submission), 2018.

Recent studies have demonstrated the possibility of inferring the evolutionary constraints acting on a protein family through homolog sequence analysis. However, the current methods are still ad hoc, and a deeper physical understanding is needed to guide more principled approaches. In a collaboration with Ned Wingreen and Anne-Flo Bitbol, I show that selection acting on any relevant physical property of a protein, e.g. the elastic energy of an important conformational change, can give rise to such a sector. We demonstrate that the main signature of these physical sectors lies in the small-eigenvalue modes of the covariance matrix of the selected sequences. This simple, generic model leads us to propose a principled method to robustly identify sectors, along with the magnitudes of mutational effects, from sequence data.

[6] S.-W. Wang, A.-F. Bitbol, N. S. Wingreen, Revealing evolutionary constraints on proteins through sequence analysis, (under review), 2018

Molecular motors play essential roles in torque generation within the cell. The conversion efficiency between chemical and mechanical energy is essential biologically. We are interested in how to make a more efficient motor. I have worked for sometime revealing the energetics and efficiency of this molecular motor, focusing on model construction directly from data. Might come back to this project in the future.