Pith. sign in

REVIEW 3 major objections 2 minor 40 references

The contact-energy cost Hamiltonian used for lattice protein structure prediction does not correlate sufficiently with structural accuracy for small peptides.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.3

2026-06-26 14:22 UTC pith:VUPPZA6B

load-bearing objection The paper reports that a standard contact-energy Hamiltonian for lattice protein folding shows weak average correlation with RMSD for small peptides, improving with instance size and more shells, based on Monte Carlo checks. the 3 major comments →

arxiv 2606.21241 v1 pith:VUPPZA6B submitted 2026-06-19 quant-ph

Assessing Cost Hamiltonian Reliability in Quantum Protein Structure Prediction

Mathieu Roget , Cedric Damour , Frederic Cadet , Jingbo Wang This is my paper
classification quant-ph
keywords quantum protein structure predictioncost Hamiltonian reliabilitycontact-energy HamiltonianRMSD correlationlattice protein modelsvariational quantum algorithmsquantum annealing
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper examines whether the contact-energy Hamiltonian that guides quantum optimization in protein structure prediction actually matches the goal of producing structures close to experimental ones. It finds that for small peptides the Hamiltonian energy landscape shows insufficient average correlation with RMSD error, so minimizing the Hamiltonian does not reliably point toward accurate folds. This matters because variational quantum algorithms and quantum annealing treat the Hamiltonian as the objective; when the proxy is misaligned, even a perfect quantum solver may return poor solutions. The authors also report that the correlation strengthens for larger instances and with additional interaction shells, estimated via Monte Carlo sampling.

Core claim

Using lattice-based quantum protein structure prediction as a case study, the contact-energy cost Hamiltonian is not sufficiently aligned with structural accuracy as measured by RMSD against experimentally determined structures. For small peptides and on average, the energy landscape of the considered cost Hamiltonian is not correlated well enough to the actual error to provide meaningful predictions. The correlation increases for larger problem instances and when more interaction shells are considered, as estimated through Monte-Carlo sampling.

What carries the argument

The contact-energy cost Hamiltonian, a simplified proxy for the true objective of structural accuracy in lattice protein models.

Load-bearing premise

That RMSD to experimentally determined structures is the appropriate ground-truth metric for judging whether the contact-energy Hamiltonian is aligned with the true task-level objective of structural accuracy.

What would settle it

A direct computation showing strong positive correlation between low contact-energy values and low RMSD for the small peptides studied would falsify the central claim.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Cost Hamiltonian relevance must be investigated independently of the quantum algorithm used.
  • Monte-Carlo sampling provides a practical way to estimate correlation for larger instances.
  • Correlation between Hamiltonian energy and RMSD error rises with problem size and with the number of interaction shells considered.

Where Pith is reading between the lines

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

  • Quantum protein prediction pipelines may require new or refined cost functions before scaling to useful sizes.
  • The same alignment check between proxy Hamiltonian and task metric could be applied to other quantum optimization problems that use simplified objectives.
  • If low correlation persists, classical pre-validation of the objective function becomes a necessary step before quantum execution.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 2 minor

Summary. The paper uses lattice-based quantum protein structure prediction as a case study to assess whether the contact-energy cost Hamiltonian aligns with structural accuracy (measured by RMSD to experimental structures). It reports that, on average for small peptides, the Hamiltonian energies show insufficient correlation with RMSD to support meaningful predictions, with the correlation estimated via Monte-Carlo sampling for larger instances and observed to increase with problem size and number of interaction shells. The work argues for evaluating cost-Hamiltonian reliability independently of the quantum algorithm employed.

Significance. If substantiated, the result would be significant for variational quantum algorithms and quantum annealing applied to optimization problems, as it demonstrates an empirical method to test Hamiltonian-task alignment and identifies a potential limitation in simplified contact-energy models for protein folding. The Monte-Carlo approach for scaling to larger instances and the observed trend with interaction shells provide concrete, falsifiable observations that could inform Hamiltonian refinement.

major comments (3)
  1. [Abstract / Results] Abstract and results: the central claim of insufficient correlation for small peptides (and its improvement for larger instances) is drawn from Monte-Carlo sampling, yet the manuscript supplies no sample sizes, error bars, exact Pearson or Spearman coefficients, or exclusion criteria, rendering it impossible to assess whether the reported lack of correlation is statistically supported.
  2. [Methods] Methods (metric definition): RMSD to experimentally determined structures is adopted as the sole ground-truth for structural accuracy without comparison to lattice-native metrics such as native-contact recovery or discrete fold-class accuracy; because lattice models define the native state via contact maps rather than continuous Cartesian RMSD, this choice leaves the mapping from low Hamiltonian energy to task success dependent on an unexamined proxy.
  3. [Results] Results (correlation trend): the reported increase in correlation for larger problem instances and additional interaction shells is presented without accompanying figures or tables that include per-instance sample statistics or confidence intervals, so the trend cannot be evaluated for robustness against sampling variability.
minor comments (2)
  1. [Methods] Notation for the contact-energy terms and interaction shells should be defined explicitly in a single location with reference to the underlying lattice model.
  2. [Methods] The Monte-Carlo sampling procedure (proposal distribution, convergence diagnostics) is described only at high level; a brief pseudocode or parameter table would improve reproducibility.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive and detailed comments. We address each major comment point by point below, indicating where revisions will be made to improve clarity and statistical rigor.

read point-by-point responses

  1. Referee: [Abstract / Results] Abstract and results: the central claim of insufficient correlation for small peptides (and its improvement for larger instances) is drawn from Monte-Carlo sampling, yet the manuscript supplies no sample sizes, error bars, exact Pearson or Spearman coefficients, or exclusion criteria, rendering it impossible to assess whether the reported lack of correlation is statistically supported.

    Authors: We agree that the current manuscript lacks sufficient detail on the Monte-Carlo procedure. In the revised version we will explicitly report the number of samples drawn for each peptide size, the exact Pearson and Spearman coefficients obtained, associated standard errors or confidence intervals, and any exclusion criteria (e.g., convergence thresholds or outlier removal). These additions will allow readers to evaluate the statistical support for the reported lack of correlation in small peptides. revision: yes

  2. Referee: [Methods] Methods (metric definition): RMSD to experimentally determined structures is adopted as the sole ground-truth for structural accuracy without comparison to lattice-native metrics such as native-contact recovery or discrete fold-class accuracy; because lattice models define the native state via contact maps rather than continuous Cartesian RMSD, this choice leaves the mapping from low Hamiltonian energy to task success dependent on an unexamined proxy.

    Authors: RMSD against experimental structures was chosen because it provides a continuous, physically interpretable measure of deviation from the true native conformation, which remains relevant even when the search space is discretized on a lattice. Nevertheless, we acknowledge that lattice-native metrics such as native-contact recovery would offer a complementary view. In the revision we will add a short discussion of native-contact recovery and, where data permit, report its correlation with the cost Hamiltonian to address the concern about an unexamined proxy. revision: partial

  3. Referee: [Results] Results (correlation trend): the reported increase in correlation for larger problem instances and additional interaction shells is presented without accompanying figures or tables that include per-instance sample statistics or confidence intervals, so the trend cannot be evaluated for robustness against sampling variability.

    Authors: We will augment the results section with supplementary figures and tables that display, for each instance size and shell count, the per-instance sample statistics, the observed correlation coefficients, and the Monte-Carlo-derived confidence intervals. This will enable direct assessment of the robustness of the reported increasing trend against sampling variability. revision: yes

Circularity Check

0 steps flagged

No significant circularity; empirical correlation study is self-contained

full rationale

The paper conducts an empirical investigation measuring the correlation between contact-energy Hamiltonian values and RMSD to experimental structures for lattice protein models. No derivation chain exists that reduces predictions or results to fitted parameters, self-citations, or ansatzes by construction. The central finding (insufficient average correlation for small peptides) is a statistical observation from Monte-Carlo sampling and direct computation against external experimental data, not a self-referential definition or renamed input. Self-citations, if present, are not load-bearing for the correlation result. The analysis is independent of the quantum algorithm and relies on verifiable external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The assessment rests on standard statistical correlation and Monte-Carlo sampling; no free parameters, ad-hoc axioms, or new entities are introduced in the abstract.

axioms (1)
  • domain assumption RMSD to experimental structures is a valid proxy for structural accuracy
    Invoked when the abstract equates low RMSD with the true task-level objective.

pith-pipeline@v0.9.1-grok · 5701 in / 1287 out tokens · 29095 ms · 2026-06-26T14:22:44.680897+00:00 · methodology

0 comments
read the original abstract

In variational quantum algorithms, QAOA, and quantum annealing, the cost Hamiltonian defines the optimization landscape explored by the quantum hardware; however, in many application-driven formulations, this Hamiltonian is a simplified proxy for the true task-level objective. Using lattice-based quantum protein structure prediction as a case study, we investigate whether the contact-energy cost Hamiltonian commonly used in this setting is sufficiently aligned with structural accuracy, as measured by RMSD against experimentally determined structures. Through this specific problem, we show the importance of studying the reliability of the cost Hamiltonian independently from the quantum approach used. This work shows that, for small peptides and on average, the energy landscape of the considered cost Hamiltonian is not correlated well enough to the actual error to provide meaningful predictions. Moreover, this correlation was estimated through Monte-Carlo sampling for larger instances. It shows an increase of said correlation for larger problem instances and when more interaction shells are considered. This investigation illustrates the meaningfulness of investigating cost Hamiltonian relevance independently from the quantum algorithm used.

Figures

Figures reproduced from arXiv: 2606.21241 by Cedric Damour, Frederic Cadet, Jingbo Wang, Mathieu Roget.

Figure 1
Figure 1. Figure 1: Full quantum scheme for solving the PSP problem (on the left) with [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Workflow of the numerical experiments done in this paper. [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Solution space for peptide 1I93 of length 11 on the tetrahedral lattice. [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: RMSD values for peptide of length ≤ 15. Parameters for the cost function are dmax = 7.8, kmin = 5 and λ = (1). average across all peptides, better than the optimal solution according to the cost Hamiltonian. This result does not indicate a failure of quantum optimization algorithms per se. Rather, it shows that, for the cost function studied here, the ideal solution of the encoded optimization problem may … view at source ↗
Figure 5
Figure 5. Figure 5: Correlation as a function of the length for proteins for all instances [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Monte-Carlo validation for instances 7CLVC, 1FDMA, 8BWFE, [PITH_FULL_IMAGE:figures/full_fig_p018_6.png] view at source ↗

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

40 extracted references · 3 canonical work pages · 1 internal anchor

  1. [1]

    Principles that govern the folding of protein chains

    Christian B Anfinsen. Principles that govern the folding of protein chains. Science, 181(4096):223–230, 1973

    1973

  2. [2]

    Non-variational quantum combinatorial optimisation

    Tavis Bennett, Lyle Noakes, and Jingbo Wang. Non-variational quantum combinatorial optimisation. In2024 IEEE International Conference on Quantum Computing and Engineering (QCE), volume 1, pages 31–41. IEEE, 2024

    2024

  3. [3]

    Bench- marking quantum heuristics: Non-variational qwoa for weighted maxcut

    Tavis Bennett, Aidan Smith, Edric Matwiejew, and Jingbo Wang. Bench- marking quantum heuristics: Non-variational qwoa for weighted maxcut. arXiv preprint arXiv:2505.24191, 2025

    work page arXiv 2025

  4. [4]

    Macmillan, 2007

    Jeremy M Berg, John L Tymoczko, and Lubert Stryer.Biochemistry (loose-leaf). Macmillan, 2007

    2007

  5. [5]

    Protein folding in the hydrophobic- hydrophilic (hp) is np-complete

    Bonnie Berger and Tom Leighton. Protein folding in the hydrophobic- hydrophilic (hp) is np-complete. InProceedings of the second annual in- ternational conference on Computational molecular biology, pages 30–39, 1998

    1998

  6. [6]

    Peptide conformational sampling using the Quantum Approximate Optimization Algorithm.npj Quantum Information, 9(1):70, July 2023

    Sami Boulebnane, Xavier Lucas, Agnes Meyder, Stanislaw Adaszewski, and Ashley Montanaro. Peptide conformational sampling using the Quantum Approximate Optimization Algorithm.npj Quantum Information, 9(1):70, July 2023. 14

    2023

  7. [7]

    Protein folding in the landscape perspective: Chevron plots and non-arrhenius kinetics.Proteins: Structure, Function, and Bioinformatics, 30(1):2–33, 1998

    Hue Sun Chan and Ken A Dill. Protein folding in the landscape perspective: Chevron plots and non-arrhenius kinetics.Proteins: Structure, Function, and Bioinformatics, 30(1):2–33, 1998

    1998

  8. [8]

    DiFilippo, Jun Qin, Daniel Blankenberg, and Omar Shehab

    Hakan Doga, Bryan Raubenolt, Fabio Cumbo, Jayadev Joshi, Frank P. DiFilippo, Jun Qin, Daniel Blankenberg, and Omar Shehab. A Perspective on Protein Structure Prediction Using Quantum Computers.Journal of Chemical Theory and Computation, 20(9):3359–3378, May 2024

    2024

  9. [9]

    Engineering protein-based ther- apeutics through structural and chemical design.Nature communications, 14(1):2411, 2023

    Sasha B Ebrahimi and Devleena Samanta. Engineering protein-based ther- apeutics through structural and chemical design.Nature communications, 14(1):2411, 2023

    2023

  10. [10]

    A Quantum Approximate Optimization Algorithm

    Edward Farhi, Jeffrey Goldstone, and Sam Gutmann. A quantum approxi- mate optimization algorithm.arXiv preprint arXiv:1411.4028, 2014

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  11. [11]

    Quantum encoding of functions and images with matrix product states.Physical Review A, 113(5):052616, 2026

    Josh Green and Jingbo Wang. Quantum encoding of functions and images with matrix product states.Physical Review A, 113(5):052616, 2026

    2026

  12. [12]

    Protein structure prediction: challenges, advances, and the shift of research paradigms.Genomics, proteomics & bioinformatics, 21(5):913–925, 2023

    Bin Huang, Lupeng Kong, Chao Wang, Fusong Ju, Qi Zhang, Jianwei Zhu, Tiansu Gong, Haicang Zhang, Chungong Yu, Wei-Mou Zheng, et al. Protein structure prediction: challenges, advances, and the shift of research paradigms.Genomics, proteomics & bioinformatics, 21(5):913–925, 2023

    2023

  13. [13]

    Highly accurate protein structure prediction with alphafold.nature, 596(7873):583–589, 2021

    John Jumper, Richard Evans, Alexander Pritzel, Tim Green, Michael Fig- urnov, Olaf Ronneberger, Kathryn Tunyasuvunakool, Russ Bates, Augustin Žídek, Anna Potapenko, et al. Highly accurate protein structure prediction with alphafold.nature, 596(7873):583–589, 2021

    2021

  14. [14]

    Quantum annealing in the transverse ising model.Physical Review E, 58(5):5355, 1998

    Tadashi Kadowaki and Hidetoshi Nishimori. Quantum annealing in the transverse ising model.Physical Review E, 58(5):5355, 1998

    1998

  15. [15]

    Are there pathways for protein folding?Journal de chimie physique, 65:44–45, 1968

    Cyrus Levinthal. Are there pathways for protein folding?Journal de chimie physique, 65:44–45, 1968

    1968

  16. [16]

    Advantages and limitations of alphafold in structural biology: Insights from recent studies.The Protein Journal, pages 1–17, 2025

    Mandarina Qing Cheng Li, Sihan Wang, Shi-Ruei Lin, Li Eric Ngok Ting, Zhi-Hong Wan, Guodong Xie, and Jane Zhang. Advantages and limitations of alphafold in structural biology: Insights from recent studies.The Protein Journal, pages 1–17, 2025

    2025

  17. [17]

    Merz Jr, Abdullah Ash Saki, Tomas Radivoyevitch, Frank DiFilippo, Jun Qin, Omar Shehab, and Daniel Blankenberg

    Rui-Hao Li, Hakan Doga, Bryan Raubenolt, Sarah Mostame, Nicholas DiSanto, Fabio Cumbo, Jayadev Joshi, Hanna Linn, Maeve Gaffney, Alexan- der Holden, Vinooth Kulkarni, Vipin Chaudhary, Kenneth M. Merz Jr, Abdullah Ash Saki, Tomas Radivoyevitch, Frank DiFilippo, Jun Qin, Omar Shehab, and Daniel Blankenberg. Quantum Algorithm for Protein Structure Prediction...

    2025

  18. [18]

    Ising formulations of many np problems.Frontiers in physics, 2:74887, 2014

    Andrew Lucas. Ising formulations of many np problems.Frontiers in physics, 2:74887, 2014. 15

    2014

  19. [19]

    A quantum walk-assisted approximate algorithm for bounded np optimisation problems.Quantum Information Processing, 18(3):1–18, 2019

    S Marsh and JB Wang. A quantum walk-assisted approximate algorithm for bounded np optimisation problems.Quantum Information Processing, 18(3):1–18, 2019

    2019

  20. [20]

    A framework for optimal quantum spatial search using alternating phase-walks.Quantum Science & Technology, 6(4):045029, 2021

    S Marsh and JB Wang. A framework for optimal quantum spatial search using alternating phase-walks.Quantum Science & Technology, 6(4):045029, 2021

    2021

  21. [21]

    Combinatorial optimization via highly efficient quantum walks.Physical Review Research, 2(2):023302, 2020

    Samuel Marsh and Jingbo Wang. Combinatorial optimization via highly efficient quantum walks.Physical Review Research, 2(2):023302, 2020

    2020

  22. [22]

    Quantum optimisation for continuous multivariable functions by a structured search.Quantum Science and Technology, 8(4):045013, 2023

    Edric Matwiejew, Jason Pye, and Jingbo B Wang. Quantum optimisation for continuous multivariable functions by a structured search.Quantum Science and Technology, 8(4):045013, 2023

    2023

  23. [23]

    Quop_mpi: A framework for parallel simulation of quantum variational algorithms.Journal of Computational Science, 62:101711, 2022

    Edric Matwiejew and Jingbo B Wang. Quop_mpi: A framework for parallel simulation of quantum variational algorithms.Journal of Computational Science, 62:101711, 2022

    2022

  24. [24]

    Quantum walk informed variational algorithm design.arXiv preprint arXiv:2406.11620, 2024

    Edric Matwiejew and Jingbo B Wang. Quantum walk informed variational algorithm design.arXiv preprint arXiv:2406.11620, 2024

    work page arXiv 2024

  25. [25]

    Jernigan

    Sanzo Miyazawa and Robert L. Jernigan. Residue – Residue Potentials with a Favorable Contact Pair Term and an Unfavorable High Packing Density Term, for Simulation and Threading.Journal of Molecular Biology, 256(3):623–644, March 1996

    1996

  26. [26]

    The protein folding problem: The role of theory.Journal of molecular biology, 433(20):167126, 2021

    Roy Nassar, Gregory L Dignon, Rostam M Razban, and Ken A Dill. The protein folding problem: The role of theory.Journal of molecular biology, 433(20):167126, 2021

    2021

  27. [27]

    Finding low-energy conformations of lattice protein models by quantum annealing.Scientific reports, 2(1):571, 2012

    Alejandro Perdomo-Ortiz, Neil Dickson, Marshall Drew-Brook, Geordie Rose, and Alán Aspuru-Guzik. Finding low-energy conformations of lattice protein models by quantum annealing.Scientific reports, 2(1):571, 2012

    2012

  28. [28]

    A variational eigenvalue solver on a photonic quantum processor.Nature communications, 5(1):4213, 2014

    Alberto Peruzzo, Jarrod McClean, Peter Shadbolt, Man-Hong Yung, Xiao- Qi Zhou, Peter J Love, Alán Aspuru-Guzik, and Jeremy L O’brien. A variational eigenvalue solver on a photonic quantum processor.Nature communications, 5(1):4213, 2014

    2014

  29. [29]

    Quantum annealing: An overview.Philosophical Transactions of the Royal Society A, 381(2241):20210417, 2023

    Atanu Rajak, Sei Suzuki, Amit Dutta, and Bikas K Chakrabarti. Quantum annealing: An overview.Philosophical Transactions of the Royal Society A, 381(2241):20210417, 2023

    2023

  30. [30]

    Barkoutsos, Stefan Woerner, and Ivano Taver- nelli

    Anton Robert, Panagiotis Kl. Barkoutsos, Stefan Woerner, and Ivano Taver- nelli. Resource-efficient quantum algorithm for protein folding.npj Quantum Information, 7(1):38, February 2021

    2021

  31. [31]

    Fast procedure for reconstruction of full-atom protein models from reduced representations.Journal of compu- tational chemistry, 29(9):1460–1465, 2008

    Piotr Rotkiewicz and Jeffrey Skolnick. Fast procedure for reconstruction of full-atom protein models from reduced representations.Journal of compu- tational chemistry, 29(9):1460–1465, 2008. 16

    2008

  32. [32]

    Alphafold and implications for intrin- sically disordered proteins.Journal of molecular biology, 433(20):167208, 2021

    Kiersten M Ruff and Rohit V Pappu. Alphafold and implications for intrin- sically disordered proteins.Journal of molecular biology, 433(20):167208, 2021

    2021

  33. [33]

    Quan- tum annealing for industry applications: Introduction and review.Reports on Progress in Physics, 85(10):104001, 2022

    Sheir Yarkoni, Elena Raponi, Thomas Bäck, and Sebastian Schmitt. Quan- tum annealing for industry applications: Introduction and review.Reports on Progress in Physics, 85(10):104001, 2022

    2022

  34. [34]

    A test of lattice protein folding algorithms

    Kaizhi Yue, Klaus M Fiebig, Paul D Thomas, Hue Sun Chan, Eugene I Shakhnovich, and Ken A Dill. A test of lattice protein folding algorithms. Proceedings of the National Academy of Sciences, 92(1):325–329, 1995

    1995

  35. [35]

    Bench- marking lie-algebraic pretraining and non-variational qwoa for the maxcut problem.arXiv e-prints, pages arXiv–2512, 2025

    Matthaus Zering, Jolyon Joyce, Tal Gurfinkel, and Jingbo Wang. Bench- marking lie-algebraic pretraining and non-variational qwoa for the maxcut problem.arXiv e-prints, pages arXiv–2512, 2025

    2025

  36. [36]

    Qdockbank: A dataset for ligand docking on protein fragments predicted on utility-level quantum computers

    Yuqi Zhang, Yuxin Yang, Cheng-Chang Lu, Weiwen Jiang, Feixiong Cheng, Bo Fang, and Qiang Guan. Qdockbank: A dataset for ligand docking on protein fragments predicted on utility-level quantum computers. InPro- ceedings of the International Conference for High Performance Computing, Networking, Storage and Analysis, pages 746–761, 2025

    2025

  37. [37]

    Efficient quantum circuits for non-unitary and unitary diagonal operators with space-time-accuracy trade-offs.ACM Transactions on Quantum Com- puting, 6(2):1–43, 2025

    Julien Zylberman, Ugo Nzongani, Andrea Simonetto, and Fabrice Debbasch. Efficient quantum circuits for non-unitary and unitary diagonal operators with space-time-accuracy trade-offs.ACM Transactions on Quantum Com- puting, 6(2):1–43, 2025. 17 A Monte Carlo validation This section presents the validation test for the Monte-Carlo estimation of the Spearman ...

    2025

  38. [38]

    , ρ100 are calculated

    100independentestimationsoftheSpearmancorrelationcoefficient( ρ1, . . . , ρ100 are calculated

  39. [39]

    The2 .5%quantile q2.5 and the97 .5%quantile q97.5 of the list( ρi)i are calculated

  40. [40]

    The results for several proteins, shows a95%confidence interval amplitude of approximately0 .05

    The amplitude of the95%confidence interval ϵ = q97.5 −q 2.5 is calculated and displayed. The results for several proteins, shows a95%confidence interval amplitude of approximately0 .05. Figure 6 shows the amplitude in function of the pro- tein’s length. The code to reproduce this experiment is available in the github repository2. Figure 6: Monte-Carlo val...