Pith. sign in

REVIEW 4 major objections 4 minor 54 references

Topological analysis reveals multiple pathways in molecular dynamics

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

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

Pith's one-line read A single learned membership function, used as a topological filter, turns molecular dynamics snapshots into a graph whose edges are distinct conformational pathways, including multiple villin folding routes.

desk verdict A clever, clearly-presented method for mapping MD pathways, but the edge rule is a geometric heuristic that needs kinetic validation before the pathway claims carry weight. read the letter →

arxiv 2412.20580 v2 pith:SYWZAIDB submitted 2024-12-29 physics.chem-ph

Luca Donati , Surahit Chewle , Dominik St. Pierre , Vijay Natarajan , Marcus Weber This is my paper · ORCID
classification physics.chem-ph
keywords moleculardynamicsconformationalpathwaysreactioncoordinateKoopmanoperatormembershipfunctionMapperalgorithmISOKANNproteinfolding
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

MoKiTo claims that one scalar membership function, learned from short molecular dynamics trajectories, can replace hand-picked reaction coordinates and still expose all the conformational pathways of a molecule. The method sorts snapshots into intervals of that function, clusters each interval by structural similarity, and connects clusters whose neighborhoods overlap; the resulting Molecular Kinetics Map comes with an energy diagram. The authors demonstrate on a two-dimensional model, 3,3-dichloroisobutene, the hexapeptide VGVAPG, and the villin headpiece subdomain that this graph reveals dominant and minor pathways, including two villin folding routes with different helix-formation orders. The payoff would be pathway identification from a representative sample and short, non-equilibrium trajectories, without a pre-defined set of collective variables.

What carries the argument

The carrying object is the $\chi$-function, the membership function for one of two metastable macro-states, obtained by applying the shift-scale iteration of ISOKANN to short trajectory data; it orders every conformational state by its progress along the slowest transition and, by construction, preserves Markovianity when used as a projection coordinate. The second mechanism is the Mapper-inspired edge rule: states are first grouped into intervals of $\chi$, clustered by common-nearest-neighbor density, and two clusters are connected only if they lie in neighboring $\chi$-intervals and their neighborhoods, defined by a threshold $r_n$ around each cluster's RMSD-averaged structure, share states. This rule turns the one-dimensional ordering into a graph whose topology is the pathway structure.

What would settle it

Generate a long unbiased trajectory for the two-dimensional model potential, count actual transitions between the four wells, and compare that crossing graph with the MoKiTo Molecular Kinetics Map; if any MKM edge connects states that never exchange trajectories, or a real crossing exists between clusters whose neighborhoods do not overlap, the central claim fails. A second check is to vary $r_n$ and the clustering parameters across a reasonable range and watch whether the number of detected pathways changes discontinuously.

Watch

Extended reading notes

Core claim

The central claim is that the membership function $\chi$ of a bi-metastable system, learned by the ISOKANN iteration, is an optimal reaction coordinate, and that a Mapper-style graph built from it recovers the full transition topology of the system. Concretely, MoKiTo subdivides the range of $\chi$ into intervals, clusters each interval with common-nearest-neighbor clustering, and assigns an edge between clusters in consecutive intervals only when their RMSD-neighborhoods overlap. The paper shows that the resulting Molecular Kinetics Maps and the energy levels computed from cluster populations reproduce known physics: two equivalent clockwise and counterclockwise rotations for 3,3-dichloroisobutene, two opening/closing routes of VGVAPG that a single end-to-end distance would hide, and a dominant and a minor villin folding pathway that differ in which helix forms first.

Load-bearing premise

The method assumes that every genuine transition runs monotonically through consecutive intervals of the membership function and that two clusters are connected whenever their averaged structures lie within a chosen RMSD threshold; if either assumption fails, the reported pathways are artifacts of the binning.

Editorial extensions

If this is right

  • If the claim holds, kinetic networks and energy diagrams can be built from short, possibly non-equilibrium trajectories plus a representative sample, avoiding microsecond simulations for initial pathway discovery.
  • Because cluster populations are read as Boltzmann weights, the relative probabilities of dominant versus minor pathways are directly available from the Molecular Kinetics Map.
  • The method can audit proposed reaction coordinates: VGVAPG's end-to-end distance correlates at 0.98 with $\chi$ yet hides two torsion-dependent routes, so a high-correlation coordinate is not necessarily sufficient.
  • For villin, the map yields concrete, testable folding hypotheses: the dominant route folds helix H3 before H2, the minor route folds H1 first, and H2 formation is the slowest step.
  • Since no collective variables are chosen in advance, the pipeline applies to systems where the relevant transition is unknown, including binding and misfolding problems.

Reading between the lines

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

  • The edge rule forbids transitions within the same $\chi$-interval; a system whose two metastable states sit at nearly equal $\chi$ values separated by a barrier would lose that edge. A natural extension is to admit within-interval edges whenever neighborhood overlap is high, and to test the difference.
  • The energy levels from cluster populations assume canonical sampling; for enhanced-sampling input such as simulated tempering, reweighting would be needed before the reported barrier heights are interpreted quantitatively.
  • A sensitivity test suggests itself: if the number or identity of pathways changes materially when the neighborhood threshold $r_n$ or the clustering parameters vary over a reasonable range, the detected multiplicity is a binning artifact, not a kinetic feature.
  • Coupling the MKM with committor estimates from the learned $\chi$ network could turn the qualitative graph into per-pathway flux probabilities, a quantitative step the paper does not take.
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 / 4 minor

Summary. The paper proposes MoKiTo, a pipeline that combines ISOKANN-derived membership functions (chi-functions) with a Mapper-inspired clustering procedure to construct Molecular Kinetics Maps and energy diagrams from molecular dynamics data. The method is applied to a two-dimensional model potential, 3,3-dichloroisobutene, the VGVAPG hexapeptide, and the villin headpiece subdomain. The authors claim that MoKiTo identifies dominant and minor conformational pathways, including multiple villin folding routes, from short trajectories without requiring a pre-defined set of collective variables. The presentation is clear and the software and data are made available, but the central pathway claims currently rest on a static geometric edge rule and lack quantitative kinetic validation.

Significance. If the pathway identifications are correct, MoKiTo would be a useful and relatively inexpensive addition to the molecular simulation analysis toolbox, particularly for systems where long unbiased simulations are impractical. The open-source implementation and the public datasets are commendable and should make the method easy to test and extend. The potential impact is real but conditional: the manuscript does not yet provide evidence that the recovered graph edges correspond to dynamical transitions, nor that the reported energy levels are statistically meaningful. The villin application, in particular, would be significant if its pathway multiplicity were quantitatively validated against committors, transition path sampling, or established Markov state model results.

major comments (4)
  1. [Methods, "Clustering and edge assignment" (Fig. 2C, Eq. 7)] The central claim that MoKiTo identifies distinct molecular pathways depends on an edge rule that is static and geometric rather than kinetic. Two clusters are connected only if they lie in consecutive chi-intervals and their rn-neighborhoods share states; no transition count, flux, or committor estimate enters the edge assignment. Because chi is a single scalar coordinate (the nc=2 case of Eq. 7), this rule also excludes by construction any pathway that is non-monotone in chi or that crosses a high barrier between structurally close clusters. The two-dimensional example cannot validate the rule, since the same overlap criterion defines the edges that are then presented as pathways; it demonstrates internal consistency, not correspondence to true transition paths. Please benchmark the edge assignments against actual short-trajectory fluxes, committor probabilities, or MSM transition probabilities in the 2D model, and report the resulting false-positive and false-negative rates for the recovered edges.
  2. [Methods, Eq. (19)] The energy levels E_Omega_i = -(1/beta) log pi_Omega_i treat normalized cluster sizes as canonical Boltzmann weights. For the villin example, the X0 states are STMD-generated and then "further equilibrated ... for 100 ps" (Results, "Villin headpiece subdomain", State space exploration and dynamics propagation), but no evidence is presented that 100 ps relaxes the STMD bias or that the sample is canonical at 300 K. Since cluster populations also depend on the arbitrary CNN parameters epsilon and theta, the relative energies in Figs. 4(D), 5(F), and 6(E) currently lack demonstrated statistical validity. Please provide convergence checks, bootstrap uncertainties, reweighting, or an explicit caveat that Eq. (19) gives only qualitative population weights.
  3. [Results, "Villin headpiece subdomain", MKM construction] The multiplicity and identities of the reported folding pathways depend on L=5, epsilon=(0.9,0.5,0.3,0.5,0.5), theta=(10,60,50,150,20), and rn=0.6. These parameters vary widely across intervals, and no sensitivity analysis is reported; the paper therefore does not establish that the blue, red, and mixed pathways in Fig. 6 survive plausible perturbations of the thresholds. Please add a stability analysis, for example by perturbing each parameter and reporting the graph edit distance or a persistence measure for the pathway decomposition across the parameter range.
  4. [Results, "Villin headpiece subdomain", Observations] The statement that the blue pathway is more likely than the red pathway, and that "mixed pathways" confirm Ref. 48, is based on the population-weighted energy diagram and visual inspection of representative structures. No committor probabilities, transition rates, or pathway fluxes are computed for villin, so the agreement with Refs. 16, 47, 48 is not demonstrated quantitatively. Please add a quantitative comparison, such as committor values for the identified pathway clusters or rates from a transition path ensemble, or explicitly limit the claims to qualitative topological descriptions.
minor comments (4)
  1. [Throughout] There are several typographical errors: "V on-Mises" should be "von Mises", "Ramachadran" should be "Ramachandran", and "the the sigmoid function" contains a duplicated article.
  2. [Fig. 5 and accompanying text] The figure caption assigns (E) to a Ramachandran plot and (F) to an energy landscape, but the text refers to "Fig. 5-(E)" as the energy diagram and "Fig. 5-(F)" as the representative structures; please harmonize the caption and the in-text references.
  3. [Results, first molecular example] The heading "33-Dichloroisobutene" is inconsistent with the chemical name "3,3-Dichloroisobutene" used in the text; please use a single consistent notation.
  4. [Background theory, Eq. (5)] The notation "x_{t+tau,m}|x_t = x" in Eq. (5) is unconventional; please clarify that the m trajectories are sampled from the conditional distribution given x_t = x, and consider writing the conditional expectation with standard probability notation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the pathway graph is built from an explicit overlap-based edge rule, but no equation or fitted parameter is renamed as a prediction, and the central claims are not forced by a self-citation chain.

full rationale

The paper's derivation chain is self-contained rather than circular. ISOKANN estimates the chi-function from short trajectories via the shift-scale iteration (Eqs. 14-16); the chi-function is then used only as an ordering filter for the Mapper-inspired clustering. The clusters are produced by CNN clustering from the MD states, and the graph edges are defined by an explicit structural-overlap criterion between clusters in consecutive chi-intervals. No fitted parameter is relabeled as a predicted pathway: the pathway set is exactly the set of paths in the thus-constructed graph. The edge rule is a modeling assumption that may be kinetically inaccurate, particularly because structural overlap of average structures is used as a proxy for transition connectivity, but this is a correctness or validity risk, not a circular reduction. The energy levels in Eq. 19 are a direct Boltzmann-population estimate; whether the STMD samples are sufficiently equilibrated to justify that estimate is a statistical question, not a circular one. References to the authors' earlier work on ISOKANN and PCCA+ supply the background algorithm and mathematical properties, but the paper does not invoke a uniqueness theorem from those works to force its pathway choices. The villin pathway conclusions are checked against independent earlier studies (Refs. 16, 47, 48), providing external, non-circular support. Overall, the central claim does not reduce to its inputs by construction.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The central claim rests on an approximate eigenfunction regression from ISOKANN, user-chosen clustering and edge thresholds, and a canonical-sampling assumption for energy estimates. No new physical entities are introduced; the MKM is a visualization and analysis object, not a new physical force, particle, or conserved quantity.

free parameters (5)
  • Number of chi-intervals L = L = 3 for 2D; L = 5 for the molecular examples
    The chi-domain is divided into regular intervals; the number is chosen by hand and affects which clusters can be connected as edges.
  • CNN clustering radius epsilon per interval = 1.0; 0.09, 0.08, 0.09, 0.07, 0.09; 0.3, 0.25, 0.2, 0.17, 0.3; 0.9, 0.5, 0.3, 0.5, 0.5
    Chosen from RMSD histograms per interval with no sensitivity analysis; cluster structure depends on these values.
  • CNN neighbor count theta per interval = 5; 5; 5; 10, 60, 50, 150, 20
    Described as varied until adequately sized clusters are found.
  • Neighborhood overlap threshold rn = 0.6; 0.05; 0.2; 0.6
    Controls which cluster pairs are assigned graph edges; chosen per system without sensitivity analysis.
  • FNN hyperparameters for ISOKANN = Layer sizes, learning rates, weight decays, epochs, and batch sizes from random search
    ISOKANN convergence and the resulting chi-function depend on these choices; the paper reports them per example.
assumptions (5)
  • domain assumption The dynamics are Markovian, ergodic, reversible, and sample a unique stationary density pi(x).
    Stated in the Background theory section as the setting for the Koopman operator.
  • domain assumption The system is effectively bimetastable for ISOKANN, so one chi-function describes the slowest process.
    All applications use the nc=2 chi-function from Eq. 7, even though the 2D potential has four minima and villin has multiple states.
  • domain assumption Cluster size is proportional to Boltzmann weight, so Eq. 19 gives energy levels.
    Requires canonical sampling of X0; questionable for STMD-derived villin states re-equilibrated for only 100 ps.
  • ad hoc to paper No transition occurs between clusters in the same chi-interval, and transitions only occur between consecutive intervals with overlapping neighborhoods.
    This edge-assignment rule is imposed in Methods and directly shapes the detected pathways.
  • ad hoc to paper CNN parameters epsilon and theta can be chosen from the RMSD histogram and by varying theta until clusters look adequate.
    No objective criterion is given; the resulting graph depends on this choice.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Topological analysis reveals multiple pathways in molecular dynamics." pith.science (2026). https://pith.science/paper/SYWZAIDB

@misc{pith2026241220580,
  author       = {Pith},
  title        = {Pith review of: Topological analysis reveals multiple pathways in molecular dynamics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/SYWZAIDB}},
  note         = {Machine review of arXiv:2412.20580}
}
read the original abstract

Molecular Dynamics simulations are essential tools for understanding the dynamic behavior of biomolecules, yet extracting meaningful molecular pathways from these simulations remains challenging due to the vast amount of generated data. In this work, we present Molecular Kinetics via Topology (MoKiTo), a novel approach that combines the ISOKANN algorithm to determine the reaction coordinate of a molecular system with a topological analysis inspired by the Mapper algorithm. Our strategy efficiently identifies and characterizes distinct molecular pathways, enabling the detection and visualization of critical conformational transitions and rare events. This method offers deeper insights into molecular mechanisms, facilitating the design of targeted interventions in drug discovery and protein engineering.

Figures

Figures reproduced from arXiv: 2412.20580 by the authors.

Figure 1
Figure 1. FIG. 1. (A) Schematic representation of an energy diagram. [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. (A) MoKiTo workflow diagram. Constructing the MKM using a three-stage procedure. (B) In the case of unknown [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. Results of the two-dimensional system. (A) Potential energy function of the two-dimensional system; (B) States of the two [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Results of 33-Dichloroisobutene molecule. (A) 33-Dichloroisobutene molecule; (B) [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Results of VGVAPG hexapeptide. (A) VGVAPG molecule; (B) [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6. Results of villin headpiece subdomain. (A) The X-ray crystal structure of villin headpiece solved at pH 6.7, green, yellow and orange [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

54 extracted references · 40 canonical work pages

  1. [1]

    Karplus \ and\ author J

    author author M. Karplus \ and\ author J. A. \ McCammon ,\ title title Molecular dynamics simulations of biomolecules , \ @noop journal journal Nature Structural Biology \ volume 9 ,\ pages 646--652 ( year 2002 ) NoStop

    work page 2002

  2. [2]

    Frenkel \ and\ author B

    author author D. Frenkel \ and\ author B. Smit ,\ @noop title Understanding Molecular Simulation: From Algorithms to Applications \ ( publisher Academic Press ,\ year 2001 ) NoStop

    work page 2001

  3. [3]

    author author M. P. \ Allen \ and\ author D. J. \ Tildesley ,\ @noop title Computer Simulation of Liquids \ ( publisher Oxford University Press ,\ year 2017 ) NoStop

    work page 2017

  4. [4]

    Laio \ and\ author M

    author author A. Laio \ and\ author M. Parrinello ,\ title title Escaping free-energy minima , \ @noop journal journal Proceedings of the National Academy of Sciences \ volume 99 ,\ pages 12562--12566 ( year 2002 a ) NoStop

    work page 2002

  5. [5]

    Dellago , author P

    author author C. Dellago , author P. G. \ Bolhuis , \ and\ author D. Chandler ,\ title title Efficient transition path sampling: Application to lennard-jones cluster rearrangements , \ @noop journal journal The Journal of Chemical Physics \ volume 108 ,\ pages 9236--9245 ( year 1998 ) NoStop

    work page 1998

  6. [6]

    Deuflhard \ and\ author M

    author author P. Deuflhard \ and\ author M. Weber ,\ title title Robust Perron cluster analysis in conformation dynamics , \ @noop journal journal Linear Algebra Appl. \ volume 398 ,\ pages 161--184 ( year 2004 ) NoStop

    work page 2004

  7. [7]

    Weber ,\ title Meshless Methods in Conformation Dynamics ,\ @noop Ph.D

    author author M. Weber ,\ title Meshless Methods in Conformation Dynamics ,\ @noop Ph.D. thesis ,\ school FU Berlin ( year 2006 ) NoStop

    work page 2006

  8. [8]

    author author R. J. \ Rabben , author S. Ray , \ and\ author M. Weber ,\ title title ISOKANN: Invariant subspaces of Koopman operators learned by a neural network , \ 10.1063/5.0015132 journal journal J. Chem. Phys. \ volume 153 ,\ pages 114109 ( year 2020 ) NoStop

    work page doi:10.1063/5.0015132

Show all 54 references
  1. [9]

    Sikorski , author E

    author author A. Sikorski , author E. Ribera Borrell , \ and\ author M. Weber ,\ title title Learning koopman eigenfunctions of stochastic diffusions with optimal importance sampling and isokann , \ 10.1063/5.0140764 journal journal Journal of Mathematical Physics \ volume 65 ...

    doi

  2. [10]

    Donati , author C

    author author L. Donati , author C. Schütte , \ and\ author M. Weber ,\ title title The kramers turnover in terms of a macro-state projection on phase space , \ 10.1080/00268976.2024.2356748 journal journal Mol. Phys. \ volume 0 ,\ pages e2356748 ( year 2024 ) NoStop

    2024

  3. [11]

    Weber \ and\ author N

    author author M. Weber \ and\ author N. Ernst ,\ title title A fuzzy-set theoretical framework for computing exit rates of rare events in potential-driven diffusion processes , \ @noop journal journal arXiv preprint arXiv:1708.00679 \ ( year 2017 ) NoStop

    2017 arXiv

  4. [12]

    Singh , author F

    author author G. Singh , author F. Memoli , \ and\ author G. Carlsson ,\ title title Topological Methods for the Analysis of High Dimensional Data Sets and 3D Object Recognition , \ in\ /10.2312/SPBG/SPBG07/091-100 booktitle Eurographics Symposium on Point-Based Graphics ,\ ed...

    doi

  5. [13]

    Lum , author G

    author author P. Lum , author G. Singh , author A. Lehman , author T. Ishkanov , author M. Vejdemo-Johansson , author M. Alagappan , author J. Carlsson , \ and\ author G. Carlsson ,\ title title Extracting insights from the shape of complex data using topology , \ @noop journa...

    2013

  6. [14]

    Vardar , author A

    author author D. Vardar , author A. Chishti , author B. Frank , author E. J. \ Luna , author A. Noegel , author S. W. \ Oh , author M. Schleicher , \ and\ author C. McKnight ,\ title title Villin-type headpiece domains show a wide range of f-actin-binding affinities , \ @noop ...

    2002

  7. [15]

    Tang , author M

    author author Y. Tang , author M. J. \ Grey , author J. McKnight , author A. G. \ Palmer III , \ and\ author D. P. \ Raleigh ,\ title title Multistate folding of the villin headpiece domain , \ @noop journal journal Journal of molecular biology \ volume 355 ,\ pages 1066--1077...

    2006

  8. [16]

    Lei , author C

    author author H. Lei , author C. Wu , author H. Liu , \ and\ author Y. Duan ,\ title title Folding free-energy landscape of villin headpiece subdomain from molecular dynamics simulations , \ 10.1073/pnas.0608432104 journal journal Proceedings of the National Academy of Science...

    doi

  9. [17]

    Lindorff-Larsen , author S

    author author K. Lindorff-Larsen , author S. Piana , author R. O. \ Dror , \ and\ author D. E. \ Shaw ,\ title title How fast-folding proteins fold , \ @noop journal journal Science \ volume 334 ,\ pages 517--520 ( year 2011 ) NoStop

    2011

  10. [18]

    Marinari \ and\ author G

    author author E. Marinari \ and\ author G. Parisi ,\ title title Simulated tempering: A new monte carlo scheme , \ 10.1209/0295-5075/19/6/002 journal journal Europhysics Letters \ volume 19 ,\ pages 451 ( year 1992 ) NoStop

    1992 doi

  11. [19]

    Sugita \ and\ author Y

    author author Y. Sugita \ and\ author Y. Okamoto ,\ title title Replica-exchange multicanonical algorithm and multicanonical replica-exchange method for simulating systems with rough energy landscape , \ @noop journal journal Chem. Phys. Lett. \ volume 329 ,\ pages 261--270 ( ...

    2000

  12. [20]

    Souaille \ and\ author B

    author author M. Souaille \ and\ author B. Roux ,\ title title Extension to the weighted histogram analysis method: combining umbrella sampling with free energy calculations , \ 10.1002/jcc.540130812 journal journal Comp. Phys- Comm. \ volume 135 ,\ pages 40–57 ( year 2001 ) NoStop

    doi

  13. [21]

    Huber , author A

    author author T. Huber , author A. Torda , \ and\ author W. van Gunsteren ,\ title title Local elevation: A method for improving the searching properties of molecular dynamics simulation , \ @noop journal journal J. Comput. Aided Mol. Des. \ volume 8 ( year 1994 ) NoStop

    1994

  14. [22]

    Laio \ and\ author M

    author author A. Laio \ and\ author M. Parrinello ,\ title title Escaping free-energy minima. \ @noop journal journal Proc. Natl. Acad. Sci. U.S.A. \ volume 99 ,\ pages 12562--6 ( year 2002 b ) NoStop

    2002

  15. [23]

    Barducci , author G

    author author A. Barducci , author G. Bussi , \ and\ author M. Parrinello ,\ title title Well-tempered metadynamics: A smoothly converging and tunable free-energy method. \ @noop journal journal Phys. Rev. Lett. \ volume 100 ,\ pages 020603 ( year 2008 ) NoStop

    2008

  16. [24]

    von Mises \ and\ author H

    author author R. von Mises \ and\ author H. Pollaczek-Geiringer ,\ title title Praktische verfahren der gleichungsaufl \"o sung . \ https://api.semanticscholar.org/CorpusID:120916484 journal journal Zamm-zeitschrift Fur Angewandte Mathematik Und Mechanik \ volume 9 ,\ pages 15...

    1929

  17. [25]

    Weber ,\ title title Implications of PCCA+ in Molecular Simulation , \ @noop journal journal Computation \ volume 6 ( year 2018 ) NoStop

    author author M. Weber ,\ title title Implications of PCCA+ in Molecular Simulation , \ @noop journal journal Computation \ volume 6 ( year 2018 ) NoStop

    2018

  18. [26]

    Keller , author X

    author author B. Keller , author X. Daura , \ and\ author W. F. \ van Gunsteren ,\ title title Comparing geometric and kinetic cluster algorithms for molecular simulation data , \ 10.1063/1.3301140 journal journal J. Chem. Phys. \ volume 132 ,\ pages 074110 ( year 2010 ) NoStop

    doi

  19. [27]

    Lemke \ and\ author B

    author author O. Lemke \ and\ author B. G. \ Keller ,\ title title Density-based cluster algorithms for the identification of core sets , \ 10.1063/1.4965440 journal journal The Journal of Chemical Physics \ volume 145 ,\ pages 164104 ( year 2016 ) NoStop

    doi

  20. [28]

    Lemke \ and\ author B

    author author O. Lemke \ and\ author B. G. \ Keller ,\ title title Common nearest neighbor clustering—a benchmark , \ 10.3390/a11020019 journal journal Algorithms \ volume 11 ( year 2018 ),\ 10.3390/a11020019 NoStop

    2018 doi

  21. [29]

    Leimkuhler \ and\ author C

    author author B. Leimkuhler \ and\ author C. Matthews ,\ @noop title Molecular Dynamics: With Deterministic and Stochastic Numerical Methods. \ ( publisher Springer ,\ address Interdisciplinary Applied Mathematics; Vol. 39 ,\ year 2015 ) NoStop

    2015

  22. [30]

    Paszke , author S

    author author A. Paszke , author S. Gross , author F. Massa , author A. Lerer , author J. Bradbury , author G. Chanan , author T. Killeen , author Z. Lin , author N. Gimelshein , author L. Antiga , author A. Desmaison , author A. Kopf , author E. Yang , author Z. DeVito , auth...

    2019

  23. [31]

    author author H. E. \ Robbins ,\ title title A stochastic approximation method , \ https://api.semanticscholar.org/CorpusID:16945044 journal journal Annals of Mathematical Statistics \ volume 22 ,\ pages 400--407 ( year 1951 ) NoStop

    1951

  24. [32]

    Eastman , author J

    author author P. Eastman , author J. Swails , author J. D. \ Chodera , author R. T. \ McGibbon , author Y. Zhao , author K. A. \ Beauchamp , author L.-P. \ Wang , author A. C. \ Simmonett , author M. P. \ Harrigan , author C. D. \ Stern , author R. P. \ Wiewiora , author B. R....

    doi

  25. [33]

    Wang , author R

    author author J. Wang , author R. M. \ Wolf , author J. W. \ Caldwell , author P. A. \ Kollman , \ and\ author D. A. \ Case ,\ title title Development and testing of a general amber force field , \ https://doi.org/10.1002/jcc.20035 journal journal J. Comp. Chem. \ volume 25 ,\...

    doi

  26. [34]

    \ Wang , author T

    author author L.-P. \ Wang , author T. J. \ Martinez , \ and\ author V. S. \ Pande ,\ title title Building force fields: An automatic, systematic, and reproducible approach , \ @noop journal journal J. Phys. Chem. Lett. \ volume 5 ,\ pages 1885--1891 ( year 2014 ) NoStop

    2014

  27. [35]

    author author J. A. \ Izaguirre , author C. R. \ Sweet , \ and\ author V. S. \ Pande ,\ title title Multiscale dynamics of macromolecules using normal mode langevin , \ @noop journal journal Pac. Symp. Biocomput. \ volume 15 ,\ pages 240--251 ( year 2010 ) NoStop

    2010

  28. [36]

    Darden , author D

    author author T. Darden , author D. York , \ and\ author L. Pedersen ,\ title title Particle mesh Ewald: An Nlog(N) method for Ewald sums in large systems , \ 10.1063/1.464397 journal journal J. Chem. Phys. \ volume 98 ,\ pages 10089--10092 ( year 1993 ) NoStop

    doi

  29. [37]

    Floquet , author S

    author author N. Floquet , author S. Héry-Huynh , author M. Dauchez , author P. Derreumaux , author A. M. \ Tamburro , \ and\ author A. J. P. \ Alix ,\ title title Structural characterization of vgvapg, an elastin-derived peptide , \ https://doi.org/10.1002/bip.20029 journal j...

    doi

  30. [38]

    Donati \ and\ author B

    author author L. Donati \ and\ author B. G. \ Keller ,\ title title Girsanov reweighting for metadynamics simulations , \ @noop journal journal J. Chem. Phys. \ volume 149 ,\ pages 072335 ( year 2018 ) NoStop

    2018

  31. [39]

    Giraldo-Barreto , author S

    author author J. Giraldo-Barreto , author S. Ortiz , author E. Thiede , author K. Palacio-Rodríguez , author B. Carpenter , author A. Barnett , \ and\ author P. Cossio ,\ title title A bayesian approach to extracting free-energy profiles from cryo-electron microscopy experimen...

    doi

  32. [40]

    author author J. A. \ Maier , author C. Martinez , author K. Kasavajhala , author L. Wickstrom , author K. E. \ Hauser , \ and\ author C. Simmerling ,\ title title ff14sb: Improving the accuracy of protein side chain and backbone parameters from ff99sb , \ @noop journal journa...

    2015

  33. [41]

    author author T. K. \ Chiu , author J. Kubelka , author R. Herbst-Irmer , author W. A. \ Eaton , author J. Hofrichter , \ and\ author D. R. \ Davies ,\ title title High-resolution x-ray crystal structures of the villin headpiece subdomain, an ultrafast folding protein , \ @noo...

    2005

  34. [42]

    author author Schrödinger, LLC ,\ https://pymol.org/2/ title The PyMOL Molecular Graphics System, Version 2.0 ,\ organization Schrödinger, LLC ( year 2015 ),\ note accessed: 2024-10-08 NoStop

    2015

  35. [43]

    \ Prinz , author H

    author author J.-H. \ Prinz , author H. Wu , author M. Sarich , author B. Keller , author M. Senne , author M. Held , author J. D. \ Chodera , author C. Sch \" u tte , \ and\ author F. No \' e ,\ title title Markov models of molecular kinetics: generation and validation. \ @no...

    2011

  36. [44]

    author author B. E. \ Husic \ and\ author V. S. \ Pande ,\ title title Markov state models: From an art to a science , \ 10.1021/jacs.7b12191 journal journal Journal of the American Chemical Society \ volume 140 ,\ pages 2386--2396 ( year 2018 ) NoStop

    doi

  37. [45]

    editor G. R. \ Bowman , editor V. S. \ Pande , \ and\ editor F. No \' e ,\ eds.,\ @noop title An Introduction to Markov State Models and Their Application to Long Timescale Molecular Simulation ,\ Vol.\ volume 797 of Advances in Experimental Medicine and Biology \ ( publisher ...

    2014

  38. [46]

    author author B. G. \ Keller , author S. Aleksic , \ and\ author L. Donati ,\ title title Markov state models in drug design , \ in\ @noop booktitle Biomolecular Simulations in Structure-based Drug Discovery ,\ editor edited by\ editor F. L. \ Gervasio \ ( publisher Wiley-Inte...

    2018

  39. [47]

    Harada \ and\ author A

    author author R. Harada \ and\ author A. Kitao ,\ title title The fast-folding mechanism of villin headpiece subdomain studied by multiscale distributed computing , \ 10.1021/ct200363h journal journal Journal of Chemical Theory and Computation \ volume 8 ,\ pages 290--299 ( ye...

    doi

  40. [48]

    Wang , author P

    author author E. Wang , author P. Tao , author J. Wang , \ and\ author Y. Xiao ,\ title title A novel folding pathway of the villin headpiece subdomain hp35 , \ 10.1039/C9CP01703H journal journal Physical Chemistry Chemical Physics \ volume 21 ( year 2019 ),\ 10.1039/C9CP01703H NoStop

    doi

  41. [49]

    merlin.mbs aapmrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs aapmrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number org...

    2010

  42. [50]

    merlin.mbs aipauth4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs aipauth4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number org...

    2010

  43. [51]

    merlin.mbs aipnum4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs aipnum4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number orga...

    2010

  44. [52]

    merlin.mbs apsrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs apsrev4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number orga...

    2010

  45. [53]

    merlin.mbs apsrmp4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked

    FUNCTION id.bst "merlin.mbs apsrmp4-1.bst 2010-07-25 4.21a (PWD, AO, DPC) hacked" ENTRY address archive archivePrefix author bookaddress booktitle chapter collaboration doi edition editor eid eprint howpublished institution isbn issn journal key language month note number orga...

    2010

  46. [54]

    write newline

    " write newline "" before.all 'output.state := FUNCTION fin.entry add.period write newline FUNCTION new.sentence output.state after.block = 'skip output.state before.all = 'skip after.sentence 'output.state := if if FUNCTION not #0 #1 if FUNCTION and 'skip pop #0 if FUNCTION o...

Pith tools

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