pith. sign in

arxiv: 2606.24283 · v1 · pith:UXVVKLSYnew · submitted 2026-06-23 · ⚛️ physics.chem-ph · cond-mat.stat-mech· physics.bio-ph

Investigating causality between principal components in protein dynamics

Debarshi Banerjee , Ali Hassanali , Alessandro Laio This is my paper

Pith reviewed 2026-06-25 22:11 UTC · model grok-4.3

classification ⚛️ physics.chem-ph cond-mat.stat-mechphysics.bio-ph
keywords protein dynamicsprincipal component analysiscausal inferencemolecular dynamicsdirected networkscollective variables
0
0 comments X

The pith

Principal components of protein motions exhibit directed causal influences missed by standard analyses.

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

The paper applies a causal discovery method to the time series of principal components extracted from molecular dynamics trajectories of two proteins. It constructs directed networks that show how one component influences another across different time scales. These directional relationships are presented as a necessary condition for causal links and are shown to be invisible to covariance-based tools such as PCA or TICA. The work argues that this approach reveals aspects of the organized dynamics among collective variables that conventional methods do not capture. A sympathetic reader would see value in a new way to describe how motion modes interact in proteins.

Core claim

By feeding the high-dimensional time series of principal components from long MD trajectories into the causal-discovery framework, the authors obtain directed networks that describe putative causal influences between components across time scales; these asymmetries are not recovered by covariance-based methods and supply information complementary to PCA and TICA.

What carries the argument

The causal-discovery framework that infers putative causal asymmetries between high-dimensional time series of principal components.

If this is right

  • Directed networks of principal-component influences can be constructed directly from MD data.
  • These networks supply directional information absent from covariance-based analyses.
  • The resulting description is complementary to both PCA and Time-lagged Independent Component Analysis.
  • The method can be used to expose previously hidden dynamical organization among collective variables.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the directions map onto functional transitions, the networks could identify which modes drive conformational change.
  • Applying the same procedure to other classes of collective variables might reveal hierarchical causal structure across biomolecular systems.
  • Cross-validation against known allosteric or folding pathways would provide an external test of the inferred links.

Load-bearing premise

The framework correctly identifies directional asymmetries in the principal-component time series that reflect real influences.

What would settle it

A controlled perturbation of one principal component in simulation or experiment that fails to produce the predicted effect on the downstream components would falsify the inferred directions.

Figures

Figures reproduced from arXiv: 2606.24283 by Alessandro Laio, Ali Hassanali, Debarshi Banerjee.

Figure 1
Figure 1. Figure 1: Ubiquitin: (a) The autocorrelation functions of the top 5 PCs, PC [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Ubiquitin: Putative causal links between PC [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: NTL9: (a) The autocorrelation functions of the top 5 PCs, PC [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: NTL9: Putative causal links between PC1 and the remaining leading principal components. Nodes represent the first five PCs, and directed edges indicate the normalized integrated Imbalance Gain, IGT, defined in Eq. 3. For each pair, the edge length and width, as well as the numerical label report the relative strength of the directional information transfer, with larger values corresponding to stronger caus… view at source ↗
read the original abstract

Principal component analysis (PCA) is widely used to characterize collective protein motions from molecular dynamics (MD) simulations. While PCA identifies the dominant modes of structural fluctuation, it does not reveal whether different principal components (PCs) causally influence each other. Here, we investigate this question using a recently introduced causal-discovery framework [Del Tatto et al, PNAS 2024], which allows to infer putative causal asymmetries between high-dimensional time series. We apply this approach to long-timescale MD trajectories of two proteins. By analyzing relationships among PCs, we construct directed networks describing how PCs influence one another across time scales. These directional relationships, whose existence is a necessary condition for the presence of a causal link, are not captured by conventional covariance-based analyses and provide information that is complementary to PCA and Time-lagged Independent Component Analysis (TICA). Our results suggest that our causal inference approach can uncover previously hidden aspects of the dynamical organization of protein motions and offer a new perspective on this very popular class of collective variables.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript applies a causal-discovery framework introduced in Del Tatto et al. (PNAS 2024) to the time series of principal components extracted from long-timescale MD trajectories of two proteins. It constructs directed networks of putative causal influences among the PCs and asserts that these directional relationships are not captured by conventional covariance-based analyses and supply information complementary to PCA and TICA.

Significance. If the Del Tatto framework can be shown to recover genuine directional asymmetries from PC time series of MD data, the approach would supply a new perspective on the dynamical organization of protein motions that is inaccessible to standard dimensionality-reduction methods.

major comments (2)
  1. [Abstract] Abstract: the central claim that the directional relationships 'are not captured by conventional covariance-based analyses and provide information that is complementary to PCA and TICA' is stated without any quantitative comparison, error estimate, or statistical test; the abstract supplies no numerical results, validation steps, or data details, leaving the claim without visible support.
  2. [Results] Application to protein trajectories: the interpretation of the reported directed networks as putative causal asymmetries rests on the assumption that the Del Tatto et al. procedure recovers ground-truth causal structure from PC time series. No controlled test on synthetic data with a known causal graph (e.g., coupled oscillators projected onto a PCA basis) is described, so it remains possible that the asymmetries arise from finite-sample effects, from the linear mixing inherent in PCA, or from violations of the method's stationarity or no-hidden-confounder assumptions that are typical in MD data.
minor comments (2)
  1. Clarify the precise definition of the time scales used when constructing the directed networks and state how many PCs were retained for each protein.
  2. Add a direct side-by-side comparison (e.g., a table or figure) showing which PC pairs exhibit directed edges that are absent from the covariance or TICA matrices.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments. We address each major point below, indicating where revisions have been made to strengthen the manuscript.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the directional relationships 'are not captured by conventional covariance-based analyses and provide information that is complementary to PCA and TICA' is stated without any quantitative comparison, error estimate, or statistical test; the abstract supplies no numerical results, validation steps, or data details, leaving the claim without visible support.

    Authors: We agree that the abstract would be strengthened by including concrete support. In the revised version we have added specific quantitative details: the number of directed edges recovered for each protein, the dominant timescales of the inferred influences, and a reference to the bootstrap procedure used to assess link stability. These elements are now cross-referenced to the Methods and Results sections while preserving the abstract's length. revision: yes

  2. Referee: [Results] Application to protein trajectories: the interpretation of the reported directed networks as putative causal asymmetries rests on the assumption that the Del Tatto et al. procedure recovers ground-truth causal structure from PC time series. No controlled test on synthetic data with a known causal graph (e.g., coupled oscillators projected onto a PCA basis) is described, so it remains possible that the asymmetries arise from finite-sample effects, from the linear mixing inherent in PCA, or from violations of the method's stationarity or no-hidden-confounder assumptions that are typical in MD data.

    Authors: The Del Tatto et al. (PNAS 2024) framework was validated on multiple synthetic benchmarks with known ground-truth graphs; we cite those results explicitly. For the MD trajectories we added stationarity diagnostics (augmented Dickey-Fuller tests on the leading PCs) and a brief discussion of possible finite-sample and hidden-confounder effects. We also inserted a paragraph noting that linear mixing by PCA is mitigated by the method's use of time-lagged conditional independence tests. A dedicated new synthetic experiment projecting coupled oscillators onto a PCA basis was not performed in the original study; we therefore treat this as a limitation and have added it to the Discussion rather than claiming full robustness. revision: partial

Circularity Check

0 steps flagged

No circularity: external method applied to new data

full rationale

The paper's central step is the application of the causal-discovery framework introduced in Del Tatto et al. (PNAS 2024) to principal-component time series extracted from two protein MD trajectories. This citation has no author overlap with the present work, and the framework is treated as an independent black-box tool whose outputs (directed networks) are not shown to reduce by construction to the input covariance structure or to any fitted parameters internal to this manuscript. No self-citation is load-bearing, no ansatz is smuggled, and no 'prediction' is statistically forced by the same data used to define the inputs. The reported directional relationships are therefore genuine outputs of an external procedure rather than tautological with the PCA coordinates themselves.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

From the abstract alone the primary unverified premise is the validity of the external causal framework when applied to protein PC time series; no free parameters, new entities, or additional axioms are introduced in the description.

axioms (1)
  • domain assumption The causal-discovery framework can infer putative causal asymmetries between high-dimensional time series of principal components.
    This premise is invoked when the authors apply the Del Tatto et al method to the PCs as stated in the abstract.

pith-pipeline@v0.9.1-grok · 5708 in / 1093 out tokens · 43357 ms · 2026-06-25T22:11:48.590152+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

137 extracted references · 43 canonical work pages · 1 internal anchor

  1. [1]

    Gorecki and J

    A. Gorecki and J. Trylska and B. Lesyng , title =. 2006 , month =. doi:10.1209/epl/i2006-10129-2 , url =

    work page doi:10.1209/epl/i2006-10129-2 2006

  2. [2]

    Extracting the Causality of Correlated Motions from Molecular Dynamics Simulations , journal =

    Hiqmet Kamberaj and Arjan. Extracting the Causality of Correlated Motions from Molecular Dynamics Simulations , journal =. 2009 , issn =. doi:https://doi.org/10.1016/j.bpj.2009.07.019 , url =

    work page doi:10.1016/j.bpj.2009.07.019 2009

  3. [3]

    PLOS Computational Biology , publisher =

    Entropy Transfer between Residue Pairs and Allostery in Proteins: Quantifying Allosteric Communication in Ubiquitin , year =. PLOS Computational Biology , publisher =. doi:10.1371/journal.pcbi.1005319 , author =

    work page doi:10.1371/journal.pcbi.1005319

  4. [4]

    Proteins: Structure, Function, and Bioinformatics , volume =

    Hacisuleyman, Aysima and Erman, Burak , title =. Proteins: Structure, Function, and Bioinformatics , volume =. doi:https://doi.org/10.1002/prot.25272 , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1002/prot.25272 , year =

    work page doi:10.1002/prot.25272

  5. [5]

    Information Transfer in Active States of Human \( _2\)-Adrenergic Receptor via Inter-Rotameric Motions of Loop Regions , volume =

    Sogunmez, Nuray and Akten, Ebru Demet , address =. Information Transfer in Active States of Human \( _2\)-Adrenergic Receptor via Inter-Rotameric Motions of Loop Regions , volume =. Applied Sciences , language =. 2022 , doi =

    2022

  6. [6]

    Physics Reports , author=

    Causality detection based on information-theoretic approaches in time series analysis , volume=. Physics Reports , author=. 2007 , month=mar, pages=. doi:10.1016/j.physrep.2006.12.004 , number=

    work page doi:10.1016/j.physrep.2006.12.004 2007

  7. [7]

    Coupling of Conformational Switches in Calcium Sensor Unraveled with Local Markov Models and Transfer Entropy , journal =

    Hempel, Tim and Plattner, Nuria and No. Coupling of Conformational Switches in Calcium Sensor Unraveled with Local Markov Models and Transfer Entropy , journal =. 2020 , doi =

    2020

  8. [8]

    Barr, Daniel and Oashi, Taiji and Burkhard, Kimberly and Lucius, Sarah and Samadani, Ramin and Zhang, Jun and Shapiro, Paul and MacKerell, Alexander D. Jr. and van der Vaart, Arjan , title =. Biochemistry , volume =. 2011 , doi =

    2011

  9. [9]

    Journal of Chemical Theory and Computation , volume =

    Qi, Yifei and Im, Wonpil , title =. Journal of Chemical Theory and Computation , volume =. 2013 , doi =

    2013

  10. [10]

    The Journal of Physical Chemistry B , volume =

    Zhang, Liqun and Centa, Thomas and Buck, Matthias , title =. The Journal of Physical Chemistry B , volume =. 2014 , doi =

    2014

  11. [11]

    Journal of Chemical Theory and Computation , volume =

    Sobieraj, Marcin and Setny, Piotr , title =. Journal of Chemical Theory and Computation , volume =. 2022 , doi =

    2022

  12. [12]

    Physical Review Letters , volume=

    Linear scaling causal discovery from high-dimensional time series by dynamical community detection , author=. Physical Review Letters , volume=. 2025 , publisher=

    2025

  13. [13]

    Causal inference for time series , volume =

    Runge, Jakob and Gerhardus, Andreas and Varando, Gherardo and Eyring, Veronika and Camps-Valls, Gustau , year =. Causal inference for time series , volume =. Nature Reviews Earth & Environment , doi =

  14. [14]

    Nature Communications , year=2022, volume=

    Jingxuan Zhu and Juexin Wang and Weiwei Han and Dong Xu , title=. Nature Communications , year=2022, volume=. doi:10.1038/s41467-022-29331- , url=

    work page doi:10.1038/s41467-022-29331- 2022

  15. [15]

    and Woolf, Thomas B

    Perilla, Juan R. and Woolf, Thomas B. , title =. The Journal of Chemical Physics , volume =. 2012 , month =. doi:10.1063/1.3702447 , url =

    work page doi:10.1063/1.3702447 2012

  16. [16]

    Scientific Reports , volume =

    Dutta, Sutapa and Ghosh, Mahua and Chakrabarti, Jaydeb , title =. Scientific Reports , volume =. 2017 , doi =

    2017

  17. [17]

    and Leahy, Daniel J

    Perilla, Juan R. and Leahy, Daniel J. and Woolf, Thomas B. , title =. Proteins: Structure, Function, and Bioinformatics , volume =. doi:https://doi.org/10.1002/prot.24257 , url =. https://onlinelibrary.wiley.com/doi/pdf/10.1002/prot.24257 , year =

    work page doi:10.1002/prot.24257

  18. [18]

    Glycobiology , volume =

    Jo, Sunhwan and Qi, Yifei and Im, Wonpil , title = ". Glycobiology , volume =. 2015 , month =. doi:10.1093/glycob/cwv083 , url =

    work page doi:10.1093/glycob/cwv083 2015

  19. [19]

    Understanding causation via correlations and linear response theory , author =. Phys. Rev. Res. , volume =. 2020 , month =. doi:10.1103/PhysRevResearch.2.043436 , url =

    work page doi:10.1103/physrevresearch.2.043436 2020

  20. [20]

    Granger Causality and Transfer Entropy Are Equivalent for Gaussian Variables , author =. Phys. Rev. Lett. , volume =. 2009 , month =. doi:10.1103/PhysRevLett.103.238701 , url =

    work page doi:10.1103/physrevlett.103.238701 2009

  21. [21]

    Contemporary Physics , volume =

    Milan Paluš , title =. Contemporary Physics , volume =. 2007 , publisher =. doi:10.1080/00107510801959206 , URL =

    work page doi:10.1080/00107510801959206 2007

  22. [22]

    Directionality of coupling from bivariate time series: How to avoid false causalities and missed connections , author =. Phys. Rev. E , volume =. 2007 , month =. doi:10.1103/PhysRevE.75.056211 , url =

    work page doi:10.1103/physreve.75.056211 2007

  23. [23]

    Information flow and causality as rigorous notions ab initio , author =. Phys. Rev. E , volume =. 2016 , month =. doi:10.1103/PhysRevE.94.052201 , url =

    work page doi:10.1103/physreve.94.052201 2016

  24. [24]

    C. W. J. Granger , journal =. Investigating Causal Relations by Econometric Models and Cross-spectral Methods , urldate =. 1969 , doi =

    1969

  25. [25]

    Granger Causality: A Review and Recent Advances

    Shojaie, Ali and Fox, Emily B. Granger Causality: A Review and Recent Advances. Annual Review of Statistics and Its Application. 2022. doi:https://doi.org/10.1146/annurev-statistics-040120-010930

    work page doi:10.1146/annurev-statistics-040120-010930 2022

  26. [26]

    Causal inference in the medical domain: a survey , volume =

    Wu, Xing and Peng, Shaoqi and Li, Jingwen and Zhang, Jian and Sun, Qun and Li, Weimin and Qian, Quan and Liu, Yue and Guo, Yike , year =. Causal inference in the medical domain: a survey , volume =. Applied Intelligence , doi =

  27. [27]

    and Greenland, Sander , title =

    Rothman, Kenneth J. and Greenland, Sander , title =. American Journal of Public Health , volume =. 2005 , doi =

    2005

  28. [28]

    Causal Inference in Sociological Research

    Gangl, Markus. Causal Inference in Sociological Research. Annual Review of Sociology. 2010. doi:https://doi.org/10.1146/annurev.soc.012809.102702

    work page doi:10.1146/annurev.soc.012809.102702 2010

  29. [29]

    2009 , isbn =

    Pearl, Judea , title =. 2009 , isbn =

    2009

  30. [30]

    Causal discovery and inference: concepts and recent methodological advances , author =. Appl. Inform. , volume =. 2016 , doi =

    2016

  31. [31]

    Social Science Computer Review , volume =

    Peter Spirtes and Clark Glymour , title =. Social Science Computer Review , volume =. 1991 , doi =

    1991

  32. [32]

    Probabilistic and Causal Inference: The Works of Judea Pearl , pages =

    Verma, TS and Pearl, Judea , title =. Probabilistic and Causal Inference: The Works of Judea Pearl , pages =. 2022 , isbn =

    2022

  33. [33]

    and Richard , editor =

    Peter Spirtes and Clark Glymour and Scheines N. and Richard , editor =. Causation, Prediction, and Search , year =

  34. [34]

    Quantifying causal coupling strength: A lag-specific measure for multivariate time series related to transfer entropy , author =. Phys. Rev. E , volume =. 2012 , month =. doi:10.1103/PhysRevE.86.061121 , url =

    work page doi:10.1103/physreve.86.061121 2012

  35. [35]

    Proceedings of the National Academy of Sciences , volume =

    Vittorio Del Tatto and Gianfranco Fortunato and Domenica Bueti and Alessandro Laio , title =. Proceedings of the National Academy of Sciences , volume =. 2024 , doi =

    2024

  36. [36]

    PNAS Nexus , volume =

    Glielmo, Aldo and Zeni, Claudio and Cheng, Bingqing and Csányi, Gábor and Laio, Alessandro , title = ". PNAS Nexus , volume =. 2022 , month =. doi:10.1093/pnasnexus/pgac039 , url =

    work page doi:10.1093/pnasnexus/pgac039 2022

  37. [37]

    Science , volume =

    George Sugihara and Robert May and Hao Ye and Chih-hao Hsieh and Ethan Deyle and Michael Fogarty and Stephan Munch , title =. Science , volume =. 2012 , doi =

    2012

  38. [38]

    Reliable detection of directional couplings using rank statistics , author =. Phys. Rev. E , volume =. 2009 , month =. doi:10.1103/PhysRevE.80.026217 , url =

    work page doi:10.1103/physreve.80.026217 2009

  39. [39]

    2004 , issn =

    Analyzing multiple nonlinear time series with extended Granger causality , journal =. 2004 , issn =. doi:https://doi.org/10.1016/j.physleta.2004.02.032 , url =

    work page doi:10.1016/j.physleta.2004.02.032 2004

  40. [40]

    Neural Computation , volume =

    Paninski, Liam , title =. Neural Computation , volume =. 2003 , month =. doi:10.1162/089976603321780272 , url =

    work page doi:10.1162/089976603321780272 2003

  41. [41]

    Partial Mutual Information for Coupling Analysis of Multivariate Time Series , author =. Phys. Rev. Lett. , volume =. 2007 , month =. doi:10.1103/PhysRevLett.99.204101 , url =

    work page doi:10.1103/physrevlett.99.204101 2007

  42. [42]

    Sample Estimate of the Entropy of a Random Vector , author =. Probl. Peredachi Inf. , volume =. 1987 , url =

    1987

  43. [43]

    and Kurths, J

    Steuer, R. and Kurths, J. and Daub, C. O. and Weise, J. and Selbig, J. , title =. Bioinformatics , volume =. 2002 , month =. doi:10.1093/bioinformatics/18.suppl_2.S231 , url =

    work page doi:10.1093/bioinformatics/18.suppl_2.s231 2002

  44. [44]

    Estimation of mutual information using kernel density estimators , author =. Phys. Rev. E , volume =. 1995 , month =. doi:10.1103/PhysRevE.52.2318 , url =

    work page doi:10.1103/physreve.52.2318 1995

  45. [45]

    Nonparametric k -nearest-neighbor entropy estimator , author =. Phys. Rev. E , volume =. 2016 , month =. doi:10.1103/PhysRevE.93.013310 , url =

    work page doi:10.1103/physreve.93.013310 2016

  46. [46]

    Estimating mutual information , author =. Phys. Rev. E , volume =. 2004 , month =. doi:10.1103/PhysRevE.69.066138 , url =

    work page doi:10.1103/physreve.69.066138 2004

  47. [47]

    Samyam Rajbhandari, Jeff Rasley, Olatunji Ruwase, and Yuxiong He

    Mutual Information between Discrete and Continuous Data Sets , year =. PLOS ONE , publisher =. doi:10.1371/journal.pone.0087357 , author =

    work page doi:10.1371/journal.pone.0087357

  48. [48]

    , title =

    Pal, Samir Kumar and Peon, Jorge and Bagchi, Biman and Zewail, Ahmed H. , title =. The Journal of Physical Chemistry B , volume =. 2002 , doi =

    2002

  49. [49]

    Escaping the Curse of Dimensionality in Estimating Multivariate Transfer Entropy , author =. Phys. Rev. Lett. , volume =. 2012 , month =. doi:10.1103/PhysRevLett.108.258701 , url =

    work page doi:10.1103/physrevlett.108.258701 2012

  50. [50]

    Measuring Information Transfer , author =. Phys. Rev. Lett. , volume =. 2000 , month =. doi:10.1103/PhysRevLett.85.461 , url =

    work page doi:10.1103/physrevlett.85.461 2000

  51. [51]

    Synchronization as adjustment of information rates: Detection from bivariate time series , author =. Phys. Rev. E , volume =. 2001 , publisher =. doi:10.1103/PhysRevE.63.046211 , url =

    work page doi:10.1103/physreve.63.046211 2001

  52. [52]

    PLOS ONE , publisher =

    Measuring Information-Transfer Delays , year =. PLOS ONE , publisher =. doi:10.1371/journal.pone.0055809 , author =

    work page doi:10.1371/journal.pone.0055809

  53. [53]

    Frontiers in Neuroinformatics , VOLUME=

    Chicharro, Daniel and Panzeri, Stefano , TITLE=. Frontiers in Neuroinformatics , VOLUME=. 2014 , URL=. doi:10.3389/fninf.2014.00064 , ISSN=

    work page doi:10.3389/fninf.2014.00064 2014

  54. [54]

    Information Flow in a Kinetic Ising Model Peaks in the Disordered Phase , author =. Phys. Rev. Lett. , volume =. 2013 , month =. doi:10.1103/PhysRevLett.111.177203 , url =

    work page doi:10.1103/physrevlett.111.177203 2013

  55. [55]

    , author=

    Synergy as a warning sign of transitions: The case of the two-dimensional Ising model. , author=. Physical review. E , year=

  56. [56]

    PLOS Computational Biology , publisher =

    Estimating Transfer Entropy in Continuous Time Between Neural Spike Trains or Other Event-Based Data , year =. PLOS Computational Biology , publisher =. doi:10.1371/journal.pcbi.1008054 , author =

    work page doi:10.1371/journal.pcbi.1008054

  57. [57]

    Directed Information Measures in Neuroscience , isbn =

    Wibral, Michael and Vicente, Raul and Lizier, Joseph , year =. Directed Information Measures in Neuroscience , isbn =. doi:10.1007/978-3-642-54474-3 , publisher=

    work page doi:10.1007/978-3-642-54474-3

  58. [58]

    and Lapish, Christopher , title =

    Timme, Nicholas M. and Lapish, Christopher , title =. 2018 , doi =. https://www.eneuro.org/content/5/3/ENEURO.0052-18.2018.full.pdf , journal =

    2018

  59. [59]

    Studies in Nonlinear Dynamics and Econometrics , doi =

    Using transfer entropy to measure information flows between financial markets , author =. Studies in Nonlinear Dynamics and Econometrics , doi =. 2013 , lastchecked =

    2013

  60. [60]

    , title = "

    Runge, J. , title = ". Chaos: An Interdisciplinary Journal of Nonlinear Science , volume =. 2018 , month =

    2018

  61. [61]

    2021 , eprint=

    Estimating Transfer Entropy via Copula Entropy , author=. 2021 , eprint=

    2021

  62. [62]

    Science , author=

    How Fast-Folding Proteins Fold , volume=. Science , author=. 2011 , month=oct, pages=. doi:10.1126/science.1208351 , number=

    work page doi:10.1126/science.1208351 2011

  63. [63]

    The Journal of Physical Chemistry B , author=

    Picosecond to Millisecond Structural Dynamics in Human Ubiquitin , volume=. The Journal of Physical Chemistry B , author=. 2016 , month=aug, pages=. doi:10.1021/acs.jpcb.6b02024 , number=

    work page doi:10.1021/acs.jpcb.6b02024 2016

  64. [64]

    2025 Towards a robust approach to infer causality from molecular dynamics simulations.The Journal of Chemical Physics162, 244105

    Towards a robust approach to infer causality from molecular dynamics simulations , volume=. The Journal of Chemical Physics , author=. 2025 , month=jun, pages=. doi:10.1063/5.0267926 , number=

    work page doi:10.1063/5.0267926 2025

  65. [65]

    Causality in Liquid Water as a Hallmark of Emergent Glassy Dynamics

    Causality in Liquid Water as a Hallmark of Emergent Glassy Dynamics , url=. arXiv , author=. doi:10.48550/arXiv.2604.19491 , note=

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2604.19491

  66. [66]

    Science , volume=

    Atomic-level characterization of the structural dynamics of proteins , author=. Science , volume=. 2010 , publisher=

    2010

  67. [67]

    Instantaneous-normal-mode theory , author=

    The short-time dynamics of molecular liquids. Instantaneous-normal-mode theory , author=. The Journal of chemical physics , volume=. 1992 , publisher=

    1992

  68. [68]

    arXiv preprint arXiv:2605.08381 , year=

    Machine learning the non-radiative decay modes in photochemical processes , author=. arXiv preprint arXiv:2605.08381 , year=

    Pith/arXiv arXiv

  69. [69]

    The Journal of Chemical Physics , author=

    Deconstructing the origins of interfacial catalysis: Why electric fields are inseparable from solvation , volume=. The Journal of Chemical Physics , author=. 2025 , month=nov, pages=. doi:10.1063/5.0288327 , number=

    work page doi:10.1063/5.0288327 2025

  70. [70]

    Journal of Chemical Theory and Computation , author=

    Do Machine-Learning Atomic Descriptors and Order Parameters Tell the Same Story? The Case of Liquid Water , volume=. Journal of Chemical Theory and Computation , author=. 2023 , month=jul, pages=. doi:10.1021/acs.jctc.2c01205 , number=

    work page doi:10.1021/acs.jctc.2c01205 2023

  71. [71]

    The Journal of Physical Chemistry Letters , author=

    Beyond Local Structures in Critical Supercooled Water through Unsupervised Learning , volume=. The Journal of Physical Chemistry Letters , author=. 2024 , month=apr, pages=. doi:10.1021/acs.jpclett.4c00383 , number=

    work page doi:10.1021/acs.jpclett.4c00383 2024

  72. [72]

    The Journal of Chemical Physics , author =

    Collective protein dynamics and nuclear spin relaxation , volume=. The Journal of Chemical Physics , author =. 1995 , month=feb, pages=. doi:10.1063/1.469213 , abstractNote=

    work page doi:10.1063/1.469213 1995

  73. [73]

    Biocomputing 2001 , pages=

    Collective reorientational motion and nuclear spin relaxation in proteins , author=. Biocomputing 2001 , pages=. 2000 , publisher=

    2001

  74. [74]

    Physical review letters , volume=

    Separation of a mixture of independent signals using time delayed correlations , author=. Physical review letters , volume=. 1994 , publisher=

    1994

  75. [75]

    The Journal of chemical physics , volume=

    Slow dynamics in protein fluctuations revealed by time-structure based independent component analysis: the case of domain motions , author=. The Journal of chemical physics , volume=. 2011 , publisher=

    2011

  76. [76]

    The Journal of chemical physics , volume=

    Identification of slow molecular order parameters for Markov model construction , author=. The Journal of chemical physics , volume=. 2013 , publisher=

    2013

  77. [77]

    Journal of chemical theory and computation , volume=

    Improvements in Markov state model construction reveal many non-native interactions in the folding of NTL9 , author=. Journal of chemical theory and computation , volume=. 2013 , publisher=

    2013

  78. [78]

    Proteins: Structure, Function, and Bioinformatics , volume=

    Collective motions in proteins: a covariance analysis of atomic fluctuations in molecular dynamics and normal mode simulations , author=. Proteins: Structure, Function, and Bioinformatics , volume=. 1991 , publisher=

    1991

  79. [79]

    Physical review letters , volume=

    Large-amplitude nonlinear motions in proteins , author=. Physical review letters , volume=. 1992 , publisher=

    1992

  80. [80]

    Proteins: Structure, Function, and Bioinformatics , volume=

    Essential dynamics of proteins , author=. Proteins: Structure, Function, and Bioinformatics , volume=. 1993 , publisher=

    1993

Showing first 80 references.