Pith. sign in

REVIEW 4 major objections 5 minor 50 references

Non-Markovianity increases transition path probability

T0 review · 4 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper shows that the standard 1/2 benchmark for transition-path probability fails once inertia or memory friction is present, with the true maximum approaching 1 for long memory.

desk verdict Core claim holds: max[p(TP|x)] can exceed 1/2 and approach 1 for long memory in the finite-mass GLE, and the analytical estimate is useful though heuristic. read the letter →

arxiv 2501.04512 v1 pith:OSW7OGBF submitted 2025-01-08 physics.chem-ph cond-mat.stat-mechphysics.bio-ph

Florian N. Brünig , Benjamin A. Dalton , Jan O. Daldrop , Roland R. Netz This is my paper · ORCID
classification physics.chem-phcond-mat.stat-mechphysics.bio-ph MSC 82C31
keywords transitionpathprobabilityreactioncoordinatequalitynon-MarkovianfrictiongeneralizedLangevinequationmemoryeffectstransmissioncoefficientinertialdynamicsproteinfolding
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper argues that the widely used benchmark value $\max[p(\mathrm{TP}|x)] = 1/2$, taken to indicate a good reaction coordinate in overdamped Markovian dynamics, is not reliable when inertia or non-Markovian memory friction is present. Using a generalized Langevin equation with a single-exponential memory kernel in a double-well potential, the authors show analytically and by simulation that the maximal transition-path probability is non-monotonic in the memory time: it dips below 1/2 at intermediate memory and approaches 1 when memory is long. The central analytical result connects this maximum to the transmission coefficient $\kappa$ via the approximate formula $\max[p(\mathrm{TP}|x)] \approx ((\kappa-1)^2+1)/(\kappa-2)^2$. If correct, this overturns a common assumption and implies that $p(\mathrm{TP}|x)$ alone cannot certify reaction-coordinate quality or Markovianity.

What carries the argument

The central object is the conditional transition-path probability $p(\mathrm{TP}|x)$, the fraction of equilibrium trajectories at position $x$ that belong to the transition-path ensemble; in the overdamped Markovian limit it equals $2\phi_A(x)\phi_B(x)$, giving a maximum of 1/2 at the barrier. The load-bearing identity is Eq. (4), $\max[p(\mathrm{TP}|x)] \approx (\kappa-1)^2+1)/(\kappa-2)^2$, which relates the peak value to the transmission coefficient $\kappa$ defined by Grote-Hynes rate theory as the ratio $k/k_\mathrm{TST}$. The derivation treats each barrier-crossing attempt as an independent trial with success probability $\kappa$, sums the geometric series of unsuccessful pairs of attempts, and thereby connects the velocity-direction committor at the barrier top to $\kappa$. This formula, together with the Grote-Hynes result $\kappa = \lambda/\omega_\mathrm{max}$, is what carries the prediction that $\max[p(\mathrm{TP}|x)] \to 1$ for long memory, $\tau \to \infty$, in the overdamped non-Markovian limit.

What would settle it

In a GLE simulation with single-exponential memory, count the number of failed barrier-crossing attempts before a successful transition starting from the barrier top with a given initial velocity, and test whether this distribution is geometric with success probability $\kappa$ from Grote-Hynes theory; a systematic mismatch would falsify the geometric-series assumption behind Eq. (4).

Watch

Extended reading notes

Core claim

The paper's central claim is that $\max[p(\mathrm{TP}|x)]$, the peak value of the conditional transition-path probability along a reaction coordinate, can exceed the overdamped Markovian benchmark value of 1/2 and can approach 1 in the presence of either inertial dynamics or long non-Markovian memory friction. The authors derive an analytical estimate, Eq. (4), expressing this maximum in terms of the Grote-Hynes transmission coefficient $\kappa$, which recovers 1 for $\kappa = 1$ and 1/2 for $\kappa = 0$. Their GLE simulations confirm that, for purely non-Markovian friction, the maximum decreases below 1/2 at intermediate memory times and then rises toward 1 as the memory time grows, so the value 1/2 is reached twice. They also show that adding even a small amount of explicit Markovian friction lowers the maximum, which reconciles their result with an earlier study that found $\max[p(\mathrm{TP}|x)] < 1/2$ in an overdamped non-Markovian setting. A protein-folding example illustrates the practical consequence: a peak near 1/2 can appear despite strongly non-Markovian dynamics.

Load-bearing premise

The analytical estimate assumes that every failed barrier-crossing attempt is independent and has the same chance of succeeding, so the number of attempts follows a simple geometric series; if attempts are correlated or the success chance changes with velocity or history, the predicted maximum $p(\mathrm{TP}|x)$ could be off, even though the qualitative simulation trend may survive.

Editorial extensions

If this is right

  • The benchmark value 1/2 for $\max[p(\mathrm{TP}|x)]$ is crossed twice in non-Markovian dynamics, once at vanishing and once at long memory, so it cannot be used as a standalone indicator of reaction-coordinate quality or Markovianity.
  • In the purely non-Markovian overdamped limit, $\max[p(\mathrm{TP}|x)]$ approaches 1 as the memory time grows, independent of the inertial time scale.
  • Adding a small fraction of Markovian friction to non-Markovian friction lowers $\max[p(\mathrm{TP}|x)]$, which explains the difference from previous results that reported values below 1/2.
  • The analytical estimate based on the Grote-Hynes transmission coefficient provides a reference curve for $\max[p(\mathrm{TP}|x)]$ in inertial or non-Markovian systems, useful for interpreting simulation data.
  • For fast-folding proteins, where memory times can be comparable to folding times, a $p(\mathrm{TP}|x)$ peak near 1/2 does not imply Markovian dynamics.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • If the central claim is right, reaction-coordinate validation protocols should either measure the memory time explicitly or use a criterion that accounts for inertia and memory, rather than relying on the absolute height of the $p(\mathrm{TP}|x)$ peak.
  • The same geometric-series logic suggests that systems with multiple memory time scales could show more complex non-monotonicity, possibly with several crossings of the 1/2 line; this is a testable extension beyond the paper's single-exponential kernel.
  • One could invert Eq. (4) in practice: given a measured $\max[p(\mathrm{TP}|x)]$ and the barrier curvature, one can infer an effective transmission coefficient and, through the Grote-Hynes relation, an effective memory time from equilibrium trajectory data alone.
  • Single-molecule experiments on bistable systems with long correlation in the environmental fluctuations might observe $\max[p(\mathrm{TP}|x)] > 1/2$; a null result would constrain the effective memory time of the experimental coordinate.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper studies the maximal value of the transition-path probability p(TP|x) for a one-dimensional reaction coordinate in a double-well potential under generalized Langevin dynamics with exponential memory friction and finite mass. It shows analytically and by simulation that max[p(TP|x)] is not bounded by the overdamped Markovian benchmark 1/2: it is nonmonotonic in the memory time, falls below 1/2 at intermediate memory in the overdamped regime, and approaches 1 for long memory or strong inertia. The authors propose an approximate relation, Eq. (4), connecting max[p(TP|x)] to the Grote-Hynes transmission coefficient kappa, test it against GLE simulations, and illustrate with the alpha3D protein that a peak near 1/2 does not imply Markovian dynamics. They conclude that p(TP|x) is not a reliable indicator of reaction-coordinate quality or Markovianity in non-Markovian systems.

Significance. If the conclusions hold, the paper corrects a common interpretation of a widely used metric: a p(TP|x) peak near 1/2 can occur even when the dynamics is strongly non-Markovian, as in the alpha3D example. The paper's strengths are the simple analytical estimate Eq. (4), the GH-based formulas, and the systematic GLE simulations covering inertial, Markovian, non-Markovian, and mixed-friction regimes. The derivations are compact, and the comparison with simulations uses the same model parameters with no fitted free parameters, which makes the qualitative message credible. The main weaknesses are the unproven independence assumption behind Eq. (4), missing statistical uncertainties in the simulation data, and some notation problems in Eq. (6); these affect the quantitative, but not necessarily the qualitative, conclusions.

major comments (4)
  1. [SM V, Eqs. (26)-(27)] The derivation of Eq. (4) assumes that each barrier recrossing attempt is an independent trial with the same success probability kappa, so that the velocity-direction commitor is kappa times sum over (1-kappa)^(2n). For non-Markovian dynamics this is an assumption, not a consequence of GH theory: kappa is a flux-weighted rate ratio, and the outcome of an attempt can be correlated with the velocity and with the history of earlier attempts through the memory kernel. The paper itself reports deviations from Eq. (4) in the overdamped intermediate-memory regime (Fig. 3B, tau_m/tau_D=0.001 near tau/tau_D ~ 1), so the formula should be framed as a heuristic estimate or supported by a direct test against simulation, e.g. by computing the velocity-direction commitor at the barrier and comparing it with kappa/(2-kappa).
  2. [Eq. (6) and SM VI] In Eq. (6), U''_max denotes the potential curvature at the barrier top, which is negative for the double-well potential; as written, kappa^2 = 1 - gamma/(tau U''_max) is larger than 1, and the associated square roots in SM VI (Eqs. (36)-(37)) are not well defined. If |U''_max| is intended, please define it explicitly and propagate the convention through Eq. (6) and SM VI. This is not merely cosmetic: Eq. (6) is the analytical basis for the tau -> infinity approach to 1 and for the tau* estimates in Fig. 3C.
  3. [Figs. 2 and 3] The simulation data are presented without error bars or confidence intervals. Since the central claims are quantitative (max[p(TP|x)] exceeds 1/2, drops below 1/2 at intermediate memory, and crosses 1/2 again at tau*), the absence of statistical uncertainties makes it difficult to assess whether the observed nonmonotonicity and the reported values of tau* are robust. Please add error bars or confidence intervals to all simulation markers, including Fig. 3C, using e.g. block averaging or bootstrapping.
  4. [SM IV and Figs. 2B/3] SM IV shows that the strict overdamped non-Markovian limit (m=0, gamma_M=0) is singular at the barrier, and the simulations therefore use tau_m/tau_D=0.001. Eq. (6) is an m -> 0 limit, but it does not by itself establish that the strict m=0 model has the same long-memory behavior. Please state explicitly whether the tau_m/tau_D -> 0 extrapolation at fixed tau/tau_D preserves max[p(TP|x)] -> 1 for tau -> infinity, or whether the finite-mass regularization is essential for the result; this would also clarify the comparison with the earlier overdamped result of Berezhkovskii and Makarov [35].
minor comments (5)
  1. [Abstract and Conclusion] The statement that the results 'disqualif[y] p(TP|x) as a criterion for reaction coordinate quality' is too strong; the benchmark remains valid in the overdamped Markovian case. Suggest 'as a universal criterion' or 'as a criterion for systems with significant inertial or non-Markovian effects'.
  2. [Page 3, text near Fig. 2] Typo: 'at there maximal value' should be 'at their maximal value'.
  3. [Eq. (6) and SM VI] The notation U''_max is used both as the signed curvature at the barrier and as its magnitude; please state the sign convention explicitly and use |U''_max| where the square root or kappa^2 formula requires it.
  4. [Fig. 2 and Fig. 3] The analytical lines for Eq. (4) are described as colored dotted lines; please ensure each curve remains identifiable in grayscale printing, for example by combining line styles with distinct marker shapes in the legends.
  5. [SM V, Eq. (28)] There appears to be an extra parenthesis in the expression <(p(TP|x=0,v))>_v; this is probably a typo for <p(TP|0,v)>_v.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Eq. (4) is an independent modeling estimate based on Grote-Hynes κ, not a fit to p(TP|x); the simulations are self-contained.

full rationale

The central analytical claim, Eq. (4), is derived in SM V from a geometric-series recrossing model in which the Grote-Hynes transmission coefficient κ plays the role of a per-attempt success probability. κ is obtained from Eqs. (5), (6), and SM VI directly from the GLE parameters (mass, friction kernel, barrier curvature), with no reference to the simulated p(TP|x) values. The mapping from κ to max[p(TP|x)] is a stated modeling assumption, not an identity, and the paper explicitly reports deviations from Eq. (4) in the overdamped intermediate-memory regime, which would not occur if the estimate had been fitted to the data. The inertial-limit formula, Eq. (3), is a direct kinetic-energy integral, and the Markovian overdamped limit, Eq. (1), is the standard Hummer relation. Self-citations (e.g., [28,30,33]) are used for context, model parameters, and the α3D illustration, but the main conclusion rests on the GLE simulations performed here and on the externally established Grote-Hynes theory [36]. No load-bearing uniqueness theorem or unverified prior result is imported from the authors' own work, and no known result is merely renamed. The non-circularity burden is therefore met; the main scientific risk is the quantitative accuracy of the geometric-series independence assumption, which is a correctness concern, not a circularity one.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The central claim relies on a single-exponential-memory GLE model, an approximate geometric-series relation between transmission coefficient and commitor, and a finite-mass regularization. These are not fitted parameters but modeling assumptions that bound the generality of the quantitative predictions.

assumptions (4)
  • domain assumption The double-well potential with a single-exponential memory kernel is a representative model for non-Markovian reaction dynamics.
    Used throughout the simulations; the paper does not test other memory kernels or potential shapes, so the quantitative universality of the non-monotonicity is not established.
  • ad hoc to paper The velocity-direction commitor can be written as a geometric series with constant success probability kappa: phi_X(0,v to X) = kappa times sum of (1-kappa)^(2n).
    SM V, Eqs. (26)-(27). Assumes independent recrossing attempts; no proof is given for non-Markovian dynamics.
  • domain assumption Grote-Hynes theory accurately gives the transmission coefficient kappa for barrier crossing in the non-Markovian and inertial regimes.
    Used to predict max[p(TP|x)] via Eq. (4); GH theory is standard but approximate away from its validity limits.
  • domain assumption Finite mass m is required for the pure non-Markovian limit gamma_M=0; the overdamped m to 0 equation becomes singular.
    SM IV, Eq. (24). This means the overdamped non-Markovian results actually use a small but finite tau_m/tau_D, so the strict overdamped limit is not covered.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Non-Markovianity increases transition path probability." pith.science (2026). https://pith.science/paper/OSW7OGBF

@misc{pith2026250104512,
  author       = {Pith},
  title        = {Pith review of: Non-Markovianity increases transition path probability},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OSW7OGBF}},
  note         = {Machine review of arXiv:2501.04512}
}
abstract

Defining low-dimensional reaction coordinates is crucial for analyzing the dynamics of complex systems and for comparison with experiments. The maximal value of the transition-path probability along the reaction coordinate $x$, $p(\mathrm{TP}|x)$, is a common estimator for reaction-coordinate quality by comparing to the theoretical maximal value of 1/2 in the overdamped Markovian limit. We show by analytical arguments and simulations that for non-Markovian dynamics $p(\mathrm{TP}|x)$ is non-monotonic as a function of the memory time and exceeds 1/2 for long memory time. This disqualifies $p(\mathrm{TP}|x)$ as a criterion for reaction coordinate quality.

Figures

Figures reproduced from arXiv: 2501.04512 by the authors.

Figure 1
Figure 1. Folding dynamics of the α3D protein for the fraction of native contacts reaction coordinate x. The free-energy profile U(x) = −kBT log[p(x)], where p(x) is the probability density of x and kBT the thermal energy, is shown as a solid blue line (left axis). The conditional probability to be on a transition path at position x, p(TP|x), is shown with bullet markers (right axis), the short horizontal gray line indicates … view at source ↗
Figure 2
Figure 2. Transition-path probability profiles, p(TP|x), between x/L = ±0.5 obtained from simulations of the GLE, Eq. (2) in a double-well potential. Analytical results are given in the overdamped limit by Eq. (1) (gray solid lines) and in the inertial limit by Eq. (3) (black broken lines). A: Results in the Markovian limit, γnM = 0, for various inertial time scales τm/τD. B: Results in the overdamped non-Markovian limit, γM … view at source ↗
Figure 3
Figure 3. A, B: Maximal values of the transition-path probability profiles, max[p(TP|x)], obtained from simulations of the GLE, Eq. (2) for γM = 0, plotted over the memory time scale τ /τD and shown for various inertial time scales τm/τD as different colored markers. The data is given for different barrier heights U0/kBT = 1.0 (A) and U0/kBT = 3.0 (B).The analytical result in the overdamped limit, Eq. (1), is indicated as a g… view at source ↗
Figures from the paper (1 more)
Figure 1
Figure 1. Figure 1: Maximal value of the transition-path probability profiles, [PITH_FULL_IMAGE:figures/full_fig_p011_1.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 45 canonical work pages

  1. [35]

    A. J. Ballard and C. Dellago, Toward the Mechanism of Ionic Dissociation in Water, J. Phys. Chem. B116, 13490 (2012)

    work page 2012

  2. [1]

    3 showingmax[p(TP|x)]→ 1 in the limitτ/τ D→∞

    Previously, max[p(TP|x)]≤ 0.5 was found for this scenario using an overdamped formulation of theGLE [35], which is in contrast to the results forγM = 0 in Fig. 3 showingmax[p(TP|x)]→ 1 in the limitτ/τ D→∞ . To elucidate on this, in Fig. 2C, profiles ofp(TP|x) are given in the overdamped non-Markovian limit (τm/τD = 0.001 and τ/τ D = 1) with gradually adde...

    work page 2020

  3. [2]

    Zijlstra, D

    N. Zijlstra, D. Nettels, R. Satija, D. E. Makarov, and B. Schuler, Transition Path Dynamics of a Dielectric Particle in a Bistable Optical Trap, Phys. Rev. Lett.125, 146001 (2020)

    work page 2020

  4. [3]

    O. K. Dudko, T. G. W. Graham, and R. B. Best, Locat- ing the Barrier for Folding of Single Molecules under an External Force, Phys. Rev. Lett.107, 208301 (2011)

    work page 2011

  5. [4]

    Truex, H

    K. Truex, H. S. Chung, J. M. Louis, and W. A. Eaton, Testing Landscape Theory for Biomolecular Processes with Single Molecule Fluorescence Spectroscopy, Phys. Rev. Lett. 115, 018101 (2015)

    work page 2015

  6. [5]

    Neupane, D

    K. Neupane, D. B. Ritchie, H. Yu, D. A. N. Foster, F. Wang, and M. T. Woodside, Transition Path Times for Nucleic Acid Folding Determined from Energy-Landscape Analysis of Single-Molecule Trajectories, Phys. Rev. Lett. 109, 068102 (2012)

    work page 2012

  7. [6]

    Frederickx, T

    R. Frederickx, T. in’t Veld, and E. Carlon, Anomalous Dynamics of DNA Hairpin Folding, Phys. Rev. Lett.112, 198102 (2014)

    work page 2014

  8. [7]

    Peters, Reaction Coordinates and Mechanistic Hypoth- esis Tests, Annu

    B. Peters, Reaction Coordinates and Mechanistic Hypoth- esis Tests, Annu. Rev. Phys. Chem.67, 669 (2016)

    work page 2016

Show all 50 references
  1. [8]

    Mehdi, D

    S. Mehdi, D. Wang, S. Pant, and P. Tiwary, Accelerating All-Atom Simulations and Gaining Mechanistic Under- standing of Biophysical Systems through State Predictive Information Bottleneck, J. Chem. Theory Comput.18, 3231 (2022)

    2022

  2. [9]

    H. Jung, R. Covino, and G. Hummer, Artificial Intel- ligence Assists Discovery of Reaction Coordinates and Mechanisms from Molecular Dynamics Simulations (2019), arXiv:1901.04595 [physics]

    2019 arXiv

  3. [10]

    M. E. Johnson and G. Hummer, Characterization of a DynamicStringMethodfortheConstructionofTransition Pathways in Molecular Reactions, J. Phys. Chem. B116, 8573 (2012)

    2012

  4. [11]

    Orland, Generating transition paths by Langevin bridges, J

    H. Orland, Generating transition paths by Langevin bridges, J. Chem. Phys.134, 174114 (2011)

    2011

  5. [12]

    W. K. Kim and R. R. Netz, The mean shape of transi- tion and first-passage paths, J. Chem. Phys.143, 224108 (2015)

    2015

  6. [13]

    J. O. Daldrop, W. K. Kim, and R. R. Netz, Transition paths are hot, Europhys. Lett.113, 18004 (2016)

    2016

  7. [14]

    Carlon, H

    E. Carlon, H. Orland, T. Sakaue, and C. Vanderzande, Effect of Memory and Active Forces on Transition Path Time Distributions, J. Phys. Chem. B122, 11186 (2018)

    2018

  8. [15]

    M. D. Louwerse and D. A. Sivak, Information Thermody- namics of the Transition-Path Ensemble, Phys. Rev. Lett. 128, 170602 (2022)

    2022

  9. [16]

    Hummer, From transition paths to transition states and rate coefficients, J

    G. Hummer, From transition paths to transition states and rate coefficients, J. Chem. Phys.120, 516 (2004)

    2004

  10. [17]

    Dellago, P

    C. Dellago, P. G. Bolhuis, and D. Chandler, Efficient transition path sampling: Application to Lennard-Jones cluster rearrangements, J. Chem. Phys.108, 9236 (1998)

    1998

  11. [18]

    P. G. Bolhuis, D. Chandler, C. Dellago, and P. L. Geissler, TRANSITION PATH SAMPLING: Throwing Ropes Over Rough Mountain Passes, in the Dark, Annu. Rev. Phys. Chem. 53, 291 (2002)

    2002

  12. [19]

    R. J. Allen, P. B. Warren, and P. R. ten Wolde, Sampling Rare Switching Events in Biochemical Networks, Phys. Rev. Lett. 94, 018104 (2005)

    2005

  13. [20]

    M. Sega, P. Faccioli, F. Pederiva, G. Garberoglio, and H. Orland, Quantitative Protein Dynamics from Dom- inant Folding Pathways, Phys. Rev. Lett.99, 118102 (2007)

    2007

  14. [21]

    Von Hansen, F

    Y. Von Hansen, F. Sedlmeier, M. Hinczewski, and R. R. Netz, Friction contribution to water-bond breakage kinet- ics, Phys. Rev. E84, 051501 (2011)

    2011

  15. [22]

    Hinczewski, Y

    M. Hinczewski, Y. von Hansen, J. Dzubiella, and R. R. Netz, How the diffusivity profile reduces the arbitrari- ness of protein folding free energies, J. Chem. Phys.132, 245103 (2010)

    2010

  16. [23]

    Neupane, A

    K. Neupane, A. P. Manuel, and M. T. Woodside, Protein folding trajectories can be described quantitatively by one- dimensional diffusion over measured energy landscapes, Nat. Phys. 12, 700 (2016)

    2016

  17. [24]

    R. B. Best and G. Hummer, Reaction coordinates and rates from transition paths, Proc. Natl. Acad. Sci. U. S. A. 102, 6732 (2005)

    2005

  18. [25]

    Peters and B

    B. Peters and B. L. Trout, Obtaining reaction coordinates by likelihood maximization, J. Chem. Phys.125, 054108 (2006)

    2006

  19. [26]

    Peters, Recent advances in transition path sampling: Accurate reaction coordinates, likelihood maximisation and diffusive barrier-crossing dynamics, Mol

    B. Peters, Recent advances in transition path sampling: Accurate reaction coordinates, likelihood maximisation and diffusive barrier-crossing dynamics, Mol. Simul.36, 1265 (2010)

    2010

  20. [27]

    Peters, P(TP|q) peak maximization: Necessary but not sufficient for reaction coordinate accuracy, Chem

    B. Peters, P(TP|q) peak maximization: Necessary but not sufficient for reaction coordinate accuracy, Chem. Phys. Lett. 494, 100 (2010)

    2010

  21. [28]

    H. S. Lee, S. H. Ahn, and E. F. Darve, The multi- dimensional generalized Langevin equation for confor- mational motion of proteins, J. Chem. Phys.150, 174113 (2019)

    2019

  22. [29]

    C. Ayaz, L. Tepper, F. N. Brünig, J. Kappler, J. O. Dal- drop, and R. R. Netz, Non-Markovian modeling of protein folding, Proc. Natl. Acad. Sci. U. S. A.118, e2023856118 (2021)

    2021

  23. [30]

    F. N. Brünig, J. O. Daldrop, and R. R. Netz, Pair- Reaction Dynamics in Water: Competition of Memory, Potential Shape, and Inertial Effects, J. Phys. Chem. B 126, 10295 (2022)

    2022

  24. [31]

    B. A. Dalton, C. Ayaz, H. Kiefer, A. Klimek, L. Tep- per, and R. R. Netz, Fast protein folding is governed by memory-dependent friction, Proc. Natl. Acad. Sci.120, e2220068120 (2023)

    2023

  25. [32]

    B. A. Dalton and R. R. Netz, pH modulates friction memory effects in protein folding (2024), arXiv:2401.12027 [physics]

    2024 arXiv

  26. [33]

    B. A. Dalton, A. Klimek, H. Kiefer, F. N. Brünig, H. Col- inet, L. Tepper, A. Abbasi, and R. R. Netz, Memory and Friction: From the Nanoscale to the Macroscale (2024), arXiv:2410.22588 [physics]

    2024 arXiv

  27. [34]

    C. Ayaz, L. Scalfi, B. A. Dalton, and R. R. Netz, Gen- eralized Langevin equation with a nonlinear potential of mean force and nonlinear memory friction from a hybrid projection scheme, Phys. Rev. E105, 054138 (2022)

    2022

  28. [36]

    A. M. Berezhkovskii and D. E. Makarov, Single-Molecule Test for Markovianity of the Dynamics along a Reaction Coordinate, J. Phys. Chem. Lett.9, 2190 (2018)

    2018

  29. [37]

    R. F. Grote and J. T. Hynes, The stable states picture of chemical reactions. II. Rate constants for condensed and gas phase reaction models, J. Chem. Phys.73, 2715 (1980)

    1980

  30. [38]

    Lindorff-Larsen, S

    K. Lindorff-Larsen, S. Piana, R. O. Dror, and D. E. Shaw, How Fast-Folding Proteins Fold, Science334, 517 (2011)

    2011

  31. [39]

    See attached Supplemental Material. 6

  32. [40]

    H. S. Chung, S. Piana-Agostinetti, D. E. Shaw, and W. A. Eaton,Structuraloriginofslowdiffusioninproteinfolding, Science 349, 1504 (2015)

    2015

  33. [41]

    J.Kappler, J.O.Daldrop, F.N.Brünig, M.D.Boehle,and R. R. Netz, Memory-induced acceleration and slowdown of barrier crossing, J. Chem. Phys.148, 014903 (2018)

    2018

  34. [42]

    Mori, Transport, collective motion, and Brownian motion, Prog

    H. Mori, Transport, collective motion, and Brownian motion, Prog. Theor. Phys.33, 423 (1965)

    1965

  35. [43]

    Zwanzig, Ensemble Method in the Theory of Irre- versibility, J

    R. Zwanzig, Ensemble Method in the Theory of Irre- versibility, J. Chem. Phys.33, 1338 (1960)

    1960

  36. [44]

    F. N. Brünig, R. R. Netz, and J. Kappler, Barrier-crossing times for different non-Markovian friction in well and barrier: A numerical study, Phys. Rev. E 106, 44133 (2022)

    2022

  37. [45]

    Hänggi, P

    P. Hänggi, P. Talkner, and M. Borkovec, Reaction-rate theory: Fifty years after Kramers, Rev. Mod. Phys.62, 251 (1990)

    1990

  38. [46]

    H. A. Kramers, Brownian motion in a field of force and the diffusion model of chemical reactions, Physica7, 284 (1940)

    1940

  39. [47]

    von Smoluchowski, Zur kinetischen Theorie der Brown- schen Molekularbewegung und der Suspensionen, Ann

    M. von Smoluchowski, Zur kinetischen Theorie der Brown- schen Molekularbewegung und der Suspensionen, Ann. Phys. 326, 756 (1906)

    1906

  40. [48]

    G. H. Weiss, First Passage Time Problems in Chemical Physics, Adv. Chem. Phys.13, 1 (1967)

    1967

  41. [49]

    Zwanzig,Nonequilibrium Statistical Mechanics (Oxford University Press, 2001)

    R. Zwanzig,Nonequilibrium Statistical Mechanics (Oxford University Press, 2001)

    2001

  42. [50]

    Onsager, Initial Recombination of Ions, Phys

    L. Onsager, Initial Recombination of Ions, Phys. Rev.54, 554 (1938). Supplemental material for: Non-Markovianity increases transition path probability Florian N. Brünig,1 Benjamin A. Dalton,1 Jan O. Daldrop,1 and Roland R. Netz1 1Department of Physics, Freie Universität Berlin...

    1938 arXiv

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.