Tethered single-legged molecular spiders on independent 1D tracks

Darko Stefanovic; David Arredondo

REVIEW 3 major objections 4 minor 32 references

Tethered single-legged molecular spiders on independent 1D tracks

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read The paper claims that teams of one-legged molecular spiders, each on its own track and connected by a tether, show a superdiffusive transient and outtravel two-legged spider teams at realistic DNAzyme rates.

desk verdict A plausible but unproven claim that tethered one-legged walkers beat two-legged teams; the result may hinge on a nonlocal hopping rule that needs a control. read the letter →

arxiv 1909.01872 v2 pith:QFV7ZYSI submitted 2019-08-27 physics.bio-ph cond-mat.stat-mechq-bio.QM

David Arredondo , Darko Stefanovic This is my paper · ORCID

classification physics.bio-phcond-mat.stat-mechq-bio.QM PACS 87.16.Nn82.39.Fk05.40.Fb02.50.Ey

keywords molecularspidersDNAzymewalkerssuperdiffusionkineticMonteCarlotetherconstraintresidencetimebiasone-dimensionaltracksboundarystate

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

The paper argues that the geometry of a molecular walker, not just its leg count, controls whether it moves superdiffusively. Specifically, it claims that tethered teams of single-legged DNAzyme spiders on separate parallel tracks exhibit a superdiffusive transient, even though an isolated one-legged spider performs only ordinary diffusion. The mechanism is that the tether keeps a detached leg close to the boundary between cleaved and uncleaved sites, while the one-legged walker's fast diffusion through the product sea lets it rebind before the team loses its grip on the boundary. If true, this gives a simpler, tunable design for synthetic walkers that can travel farther and sustain directed motion longer at realistic enzyme cleavage rates.

What carries the argument

The central object is a team of $w$ one-legged walkers on parallel independent 1D tracks, coupled by a tether of maximum length $d$. The walker's state is reduced to a residence-time bias: a leg on an uncleaved substrate site hops at rate $r<1$, and a leg on a cleaved product site hops at rate $1$. The argument uses the boundary-state/diffusive-state decomposition: a boundary period (B) begins when at least one leg is cleaving a substrate, and a diffusive period (D) is the random walk through the product sea; the team takes a step when the average leftmost substrate position increases by one. The analytical $r\to 0$ case reduces the team to a Markov chain on the integer difference between leftmost substrates per track, with survival probability $\Pi$, and yields expected steps $\langle S_{n=1,w=2}(r\to0)\rangle = 2d+1$ for two one-legged spiders. The tether is the mechanism that converts the diffusive motion of detached legs into renewed boundary contact.

What would settle it

Run the same kinetic Monte Carlo simulation with nearest-neighbor hopping for detached legs, keeping $r=0.1$, the same tether lengths $d$, and the same team sizes $w$, and compare expected boundary steps and the exponent $\alpha$ in $\langle x^2\rangle = 2Dt^\alpha$; if one-legged teams no longer show a longer superdiffusive transient than two-legged teams, the paper's conclusion is an artifact of the instantaneous-rebinding rule.

Watch

Extended reading notes

Core claim

The paper claims that a team of single-legged molecular spiders, each confined to its own one-dimensional track and connected by a tether that limits how far any two legs can separate, undergoes transient superdiffusive motion even though an isolated one-legged spider does not. At the realistic cleavage rate $r=0.1$, the one-legged teams make more team steps per boundary period and diffuse faster through the product sea than two-legged tethered teams, so they sustain superdiffusion for longer and travel farther. The tether keeps a leg that has detached from the boundary close enough to rebind before the team loses its last substrate-bound member, and the one-legged walkers' faster diffusion in the visited region makes this rebinding more likely.

Load-bearing premise

The argument's comparison at $r=0.1$ rests on a simulation rule in which a detached leg instantly jumps to any site allowed by the tether with equal probability, including its own site, instead of hopping neighbor-by-neighbor; if that rule is wrong, the central advantage of one-legged teams may disappear.

Editorial extensions

If this is right

At realistic cleavage rates, tethering single-legged walkers into teams produces a superdiffusive transient that an isolated one-legged walker does not have.
One-legged teams make more team steps per boundary period than two-legged teams, and exceed them in mean displacement at long times.
Longer tethers help up to an optimum: short enough to keep detached legs near the boundary, long enough not to impede forward progress.
The same residence-time-bias plus tether mechanism applies to any stochastic walker with slower first visits, not just DNAzyme spiders.

Reading between the lines

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

A testable extension would replace the instantaneous-rebinding rule with nearest-neighbor hopping; the one-legged advantage at $r=0.1$ is likely sensitive to this choice.
The paper's equivalence of a one-legged team to a multi-legged spider on independent tracks suggests that tether design can substitute for leg count in future synthetic walkers.
The analytical result that a two-spider one-legged team has survival probability exactly $1/2$ implies the advantage is entirely a diffusion-and-tether effect, which a direct measurement of rebinding probabilities could test.
Varying track dimensionality, such as 2D surfaces or branched tracks, could change the optimal tether length and the size of the superdiffusive window.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Short version: this paper adds a new configuration to the spider-walker literature—teams of one-legged walkers on parallel tracks, connected by a tether—and reports a simulation result that at realistic cleavage rates these teams spend longer in boundary periods and superdiffuse further than two-legged teams. The r→0 analysis shows the opposite, which the authors acknowledge; the advantage appears only at finite r. That is a testable claim, and the setup is clean enough to reimplement. But the central comparison leans on a nonlocal rebinding rule in Sec. III: a detached leg can jump instantly to any site within the tether constraint. In the r→0 limit that's equivalent to infinitely fast diffusion, but at r=0.1 it's not. A leg that leaves the boundary can land back on it in one move, which shortens diffusive excursions and inflates boundary-period lengths. That bias likely acts more strongly for one-legged teams (no second leg to slow diffusion), so the 'one-legged advantage' could be an artifact. No local-hopping control, no sensitivity analysis, and no error bars on the simulated MSD or α are reported.

Also, Eq. (5) is inconsistent with Eq. (9) and Table I: for Π=0.5 it gives d+1, but the correct result is 2d+1. Probably a typo, but it needs fixing. The α>1.1 threshold is arbitrary, though it doesn't change the qualitative comparison. There is no code or data for independent checks.

What the paper does well: it isolates the tether effect from the multi-leg effect, clearly inherits Rank et al.'s framework, and honestly notes that the analytic limit favors two-legged teams. The model is described well enough to be reimplemented. The idea is worth pursuing; the execution needs another pass.

My recommendation: send it to peer review, and ask for a revision that (1) fixes the formula inconsistency, (2) runs a local-hopping control or provides molecular-scale justification for the jump rule, (3) reports error bars and number of runs, and (4) releases code. If the nonlocal rule turns out to be the reason for the one-legged advantage, the paper's main claim fails; if the result survives local hopping, it's a nice design rule.

Referee Report

3 major / 4 minor

Summary. The paper studies teams of single-legged molecular spiders on independent parallel 1D tracks connected by a flexible tether. The authors extend the r→0 analytical framework of Rank et al. to one-legged teams and compare the expected number of steps per boundary period with two-legged teams. They then use kinetic Monte Carlo simulations at a realistic cleavage rate r=0.1 to argue that one-legged teams outperform two-legged teams: they diffuse faster through the product sea, remain in the boundary state longer, and exhibit a longer superdiffusive transient. The central claims are that tethering can induce superdiffusion in one-legged walkers and that one-legged teams are superior to two-legged teams at realistic parameters.

Significance. If the finite-rate results are correct, the paper identifies a design principle—tethering one-legged walkers—that improves both processivity and superdiffusive persistence in synthetic DNAzyme walkers, which would be a valuable contribution to the molecular-spider literature. The r→0 analysis for one-legged teams is a natural and potentially useful extension of Rank et al., and the symmetry argument giving Π=0.5 is clean. However, the central finite-r conclusion rests on a kinetic Monte Carlo update rule that is not validated against local hopping, and the analytical section contains an algebraic inconsistency. The significance is therefore conditional on resolving these issues.

major comments (3)

[Section III and Figures 4–7] The kinetic Monte Carlo model lets a detached leg 'instantly move to any site within its constraints (including the same site)' with uniform probability. This non-nearest-neighbor jump rule does not correspond to physical diffusion of a DNAzyme leg on a track and strongly affects the central comparison: a leg at distance ℓ from the boundary can return in one move under this rule, whereas local hopping would require O(ℓ²) steps. Because the one-legged advantage at r=0.1 in Figures 4–7 is driven by the rate at which detached legs re-enter the boundary, the claimed superiority may be an artifact of this choice. The paper provides no molecular-scale justification and no sensitivity analysis for the jump kernel. I request either a local-hopping simulation (with the tether enforced as a connectivity constraint) or a systematic comparison of jump kernels, and a discussion of how the results depend on this modeling assumption.
[Section II, Eqs. (5) and (9), Table I] Equation (5) states ⟨S_{n=1}(r→0)⟩ = d + Π/(1−Π), which for Π=0.5 gives d+1. Equation (9) and Table I instead give 2d+1 (e.g., d=2 gives 5, not 3). The correct expression is (d+Π)/(1−Π), which reduces to 2d+1 when Π=0.5. The missing parentheses in Eq. (5) are not a harmless typo because they change the r→0 prediction and the interpretation of the analytical comparison; the equation should be corrected and the surrounding derivation checked.
[Section IV, Figures 4–7] The simulation results are presented as point estimates without error bars, confidence intervals, or any other measure of statistical uncertainty. This is particularly relevant for the headline claims that one-legged teams have 'greater' expected steps per boundary period (Figure 4) and 'longer' superdiffusive transients (Figure 7), where the differences between one- and two-legged teams are the central result. Reporting standard errors or confidence bands, at least for the main parameter combinations, would allow the reader to judge whether the observed differences are significant rather than sampling noise.

minor comments (4)

[Section IV.A, first paragraph] The text says 'We ue the valuer = 0.1' and 'as it is furthest from r→0'; 'ue' should be 'use' and 'valuer' should be 'value'.
[Section IV.B, first paragraph] The sentence 'The distribution of displacements for 10^4 simulations is shown for the same subset of parameters in Figure 5' appears to reference the wrong figure: Figure 5 shows the distribution of steps per boundary period, while the displacement distributions are in Figure 6.
[Figure 6 caption] The caption repeats 'Curves were generated using the kernel density estimate in seaborn [29]' for both panels; the wording could be streamlined, and it would be helpful to state how many samples underlie each curve.
[Section II, Eq. (3)] The state-space diagram in Eq. (3) is not fully explained in the text; in particular, the meaning of the transitions labeled '1/2' versus 'Π' should be stated explicitly, since the reader must otherwise reconstruct the convention from Rank et al.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the analytic results follow from inherited Markov-chain state spaces and symmetry, and the finite-r results are simulation outputs under a stated hop rule, not fitted predictions.

full rationale

The paper's derivation chain is self-contained with respect to its target claims. The r->0 analytic results are not fitted to the later predictions: the two-legged survival probability and expected steps are taken from Rank et al. (Eqs. 14, 15, 17), while the one-legged survival probability is obtained 'by symmetry' as 1/2 and the one-legged expected steps follow from the analogous state space in Eq. (3), giving Eq. (9) directly. No fitted parameter is renamed as a prediction; the finite-r central comparisons (Figures 4-7) are kinetic Monte Carlo outputs. The value r = 0.1 is chosen from an experimental dwell-time range reported in Pei et al., an external measurement, not from the target superdiffusion results. Self-citations to Semenov et al. provide context (B/D metastates, a diffusion constant) but are not load-bearing uniqueness claims; even if those citations were ignored, the simulations stand on their own. The nonlocal 'instantly move to any site within its constraints' rule in Section III is a stated modeling assumption whose physical realism could be questioned, but it is an input to the simulation, not a quantity derived from the system's outputs. There is no equation in the paper that reduces to its own input, and no statistical forcing of the predicted one-legged advantage. Accordingly, the paper exhibits no circularity of the kind that would raise the score.

Assumptions & free parameters 2 free parameters · 5 assumptions · 0 invented entities

The model relies on a small set of domain assumptions inherited from the molecular-spider literature: a single-parameter residence-time bias, instantaneous product-site diffusion in the r→0 limit, nonlocal rebinding in the simulation, and transfer of Rank et al.'s state-space methods. No new physical entities are postulated; the one-legged tethered team is a new configuration of existing walker and tether models.

free parameters (2)

r (substrate hopping rate) = 0.1
Chosen as a realistic DNAzyme value from the experimental dwell-time ratio range (0.05,0.1) cited to Pei et al.; it is a control parameter, not fitted to the target results.
superdiffusion threshold alpha = 1.1
Arbitrary classification threshold used to define significant superdiffusive motion in Fig. 7; not derived from data or theory.

assumptions (5)

domain assumption Hopping rates on substrate vs product are captured by a single parameter r, with r=1 on product sites and r<1 on substrate sites.
Abstracts all DNAzyme chemistry into one rate; introduced in Section I and used throughout; supported by previous spider models [20,21,23].
domain assumption In the r→0 limit, movement on visited sites is effectively instantaneous, so a spider that completes a cleavage instantly diffuses to the next substrate within tether range.
Used in Section II to reduce the team to the finite Markov chains in Eqs. (2) and (3); inherited from Rank et al.
domain assumption A detached leg can move to any site within the tether and no-overtaking constraints with uniform probability, not just to nearest neighbors.
Stated in Section III and Fig. 1; load-bearing for the kinetic Monte Carlo return-to-boundary dynamics and not separately justified by molecular data.
standard math The state-space and survival-probability machinery of Rank et al. for two-legged tethered teams carries over to one-legged teams.
Section II imports Rank's Eqs. (14), (15), and (17) and constructs the analogous one-legged state space Eq. (3); the carry-over is asserted rather than fully re-derived.
domain assumption r=0.1 is representative of experimental DNAzyme dwell-time ratios.
Section III states the experimental range (0.05,0.1) from Pei et al. and uses r=0.1 for simulations; results for other non-zero r values are not shown.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Tethered single-legged molecular spiders on independent 1D tracks." pith.science (2026). https://pith.science/paper/QFV7ZYSI

@misc{pith2026190901872,
  author       = {Pith},
  title        = {Pith review of: Tethered single-legged molecular spiders on independent 1D tracks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QFV7ZYSI}},
  note         = {Machine review of arXiv:1909.01872}
}

read the original abstract

We study the motion of random walkers with residence time bias between first and subsequent visits to a site, as a model for synthetic molecular walkers composed of coupled DNAzyme legs known as molecular spiders. The mechanism of the transient superdiffusion has been explained via the emergence of a boundary between the new and the previously visited sites, and the tendency of the multi-legged spider to cling to this boundary, provided residence time for a first visit to a site is longer than for subsequent visits. Using both kinetic Monte Carlo simulation and an analytical approach, we model a system that consists of single-legged walkers, each on its own one-dimensional track, connected by a "leash", i.e., a kinematic constraint that no two spiders can be more than a certain distance apart. Even though a single one-legged walker does not at all exhibit directional, superdiffusive motion, we find that a team of one-legged walkers on parallel tracks, connected by a flexible tether, does enjoy a superdiffusive transient. Furthermore, the one-legged walker teams exhibit a greater expected number of steps per boundary period and are able to diffuse more quickly through the product sea than two-legged walkers, which leads to longer periods of superdiffusion.

Figures

Figures reproduced from arXiv: 1909.01872 by the authors.

**Figure 2.** FIG. 2. The entries of the matrices above are the possible [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** FIG. 3. The distribution of substrates was recorded at the [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (3 more)

**Figure 5.** Figure 5: FIG. 5. Probability density for number of steps per bound [PITH_FULL_IMAGE:figures/full_fig_p005_5.png]

**Figure 6.** Figure 6: FIG. 6. Probability Density of spider positions at times [PITH_FULL_IMAGE:figures/full_fig_p006_6.png]

**Figure 7.** Figure 7: FIG. 7. Mean Square Displacement and finite difference ap [PITH_FULL_IMAGE:figures/full_fig_p007_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 31 canonical work pages

[1]

Schliwa and G

M. Schliwa and G. Woehlke, Nature 422, 759 (2003)

2003
[2]

The distribution of 6 displacements for 10 4 simulations is shown for the same subset of parameters in Figure 5

The biased random walk will have a positive mean displacement as a function of time. The distribution of 6 displacements for 10 4 simulations is shown for the same subset of parameters in Figure 5. The mean square displacement (MSD) in one- dimensional space is deﬁned as ⟨x2⟩ = 2Dtα, (10) where D is the diﬀusion constant, and is aﬀected by the geometry of...

work page
[3]

Ait-Haddou and W

R. Ait-Haddou and W. Herzog, Cell Biochemistry and Biophysics 38, 191 (2003)

work page 2003
[4]

R. D. Vale, Cell 112, 467 (2003)

work page 2003
[5]

M. A. Geeves, Biopolymers 105, 483 (2016)

work page 2016
[6]

Krendel and M

M. Krendel and M. S. Mooseker, Physiology 20, 239 (2005)

work page 2005
[7]

R. A. Muscat, J. Bath, and A. J. Turberﬁeld, Small 8, 3593 (2012)

work page 2012
[8]

The special case of r→0 would realistically not allow motion in the positive direction, but permits simpliﬁca- tions in the state space of the system

Note that this deﬁnition implies that after the spider’s forward leg has cleaved the substrate site, it can take any number of half-steps back and forth from its original position before completing a whole step in either direction. The special case of r→0 would realistically not allow motion in the positive direction, but permits simpliﬁca- tions in the s...

work page

Show all 32 references

[9]

Omabegho, R

T. Omabegho, R. Sha, and N. C. Seeman, Science 324, 67 (2009)

2009
[10]

A. J. Thubagere, W. Li, R. F. Johnson, Z. Chen, S. Doroudi, Y. L. Lee, G. Izatt, S. Wittman, N. Srini- vas, D. Woods, E. Winfree, and L. Qian, Science 357, eaan6558 (2017)

2017
[11]

J. Bath, S. J. Green, and A. J. Turberﬁeld, Angewandte Chemie International Edition 44, 4358 (2005)

2005
[12]

S. F. J. Wickham, M. Endo, Y. Katsuda, K. Hidaka, J. Bath, H. Sugiyama, and A. J. Turberﬁeld, Nature Nanotechnology 6, 166 (2011)

2011
[13]

S. F. J. Wickham, J. Bath, Y. Katsuda, M. Endo, K. Hi- daka, H. Sugiyama, and A. J. Turberﬁeld, Nature Nan- otechnology 7, 169 (2012)

2012
[14]

X. Yang, Y. Tang, S. D. Mason, J. Chen, and F. Li, ACS Nano 10, 2324 (2016)

2016
[15]

Y. Ji, L. Zhang, L. Zhu, J. Lei, J. Wu, and H. Ju, Biosen- sors and Bioelectronics 96, 201 (2017)

2017
[16]

Y. Tian, Y. He, Y. Chen, and C. Mao, Angewandte Chemie International Edition 44, 4355 (2005)

2005
[17]

K. Lund, A. J. Manzo, N. Dabby, N. Michelotti, A. Johnson-Buck, J. Nangreave, S. Taylor, R. Pei, M. N. Stojanovic, N. G. Walter, E. Winfree, and H. Yan, Na- ture 465, 206 (2010)

2010
[18]

T.-G. Cha, J. Pan, H. Chen, J. Salgado, X. Li, C. Mao, and J. H. Choi, Nature Nanotechnology 9, 39 (2014)

2014
[19]

T.-G. Cha, J. Pan, H. Chen, H. N. Robinson, X. Li, C. Mao, and J. H. Choi, Journal of the American Chem- ical Society 137, 9429 (2015)

2015
[20]

S. Cai, M. Chen, M. Liu, W. He, Z. Liu, D. Wu, Y. Xia, H. Yang, and J. Chen, Biosensors and Bioelectronics 85, 184 (2016)

2016
[21]

K. Yang, H. Wang, N. Ma, M. Zeng, H. Luo, and D. He, ACS Applied Materials & Interfaces 10, 44546 (2018)

2018
[22]

M. Rank, L. Reese, and E. Frey, Physical Review E 87, 032706 (2013)

2013
[23]

Antal and P

T. Antal and P. L. Krapivsky, Physical review. E, Sta- tistical, nonlinear, and soft matter physics 76, 021121 (2007)

2007
[24]

Semenov, M

O. Semenov, M. J. Olah, and D. Stefanovic, Natural Computing 12, 259 (2013)

2013
[25]

Semenov, M

O. Semenov, M. J. Olah, and D. Stefanovic, Physical Review E 83, 021117 (2011)

2011
[26]

Semenov, D

O. Semenov, D. Mohr, and D. Stefanovic, Physical Re- view E 88, 012724 (2013)

2013
[27]

S. W. Santoro and G. F. Joyce, Biochemistry 37, 4262 (1998)

1998
[28]

R. Pei, S. K. Taylor, D. Stefanovic, S. Rudchenko, T. E. Mitchell, and M. N. Stojanovic, Journal of the American Chemical Society 128, 1269312699 (2006)

2006
[29]

M. J. Olah and D. Stefanovic, Physical Review E 87, 062713 (2013)

2013
[30]

Stefanovic, M

D. Stefanovic, M. N. Stojanovic, M. J. Olah, and O. Se- menov, in 12th European Conference on Artiﬁcial Life (2013)

2013
[31]

mwaskom/seaborn: v0.9.0 (july 2018),

M. Waskom, O. Botvinnik, D. O’Kane, P. Hobson, J. Os- tblom, S. Lukauskas, D. C. Gemperline, T. Augspurger, Y. Halchenko, J. B. Cole, J. Warmenhoven, J. de Ruiter, C. Pye, S. Hoyer, J. Vanderplas, S. Villalba, G. Kunter, E. Quintero, P. Bachant, M. Martin, K. Meyer, A. Miles, ...

2018
[32]

Press, S

W. Press, S. A. Teukolsky, W. T. Vetterling, and B. P. Flannery, Numerical Recipes in C++ (Cambridge Uni- versity Press, New York, 2002)

2002

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Schliwa and G

M. Schliwa and G. Woehlke, Nature 422, 759 (2003)

2003

[2] [2]

The distribution of 6 displacements for 10 4 simulations is shown for the same subset of parameters in Figure 5

The biased random walk will have a positive mean displacement as a function of time. The distribution of 6 displacements for 10 4 simulations is shown for the same subset of parameters in Figure 5. The mean square displacement (MSD) in one- dimensional space is deﬁned as ⟨x2⟩ = 2Dtα, (10) where D is the diﬀusion constant, and is aﬀected by the geometry of...

work page

[3] [3]

Ait-Haddou and W

R. Ait-Haddou and W. Herzog, Cell Biochemistry and Biophysics 38, 191 (2003)

work page 2003

[4] [4]

R. D. Vale, Cell 112, 467 (2003)

work page 2003

[5] [5]

M. A. Geeves, Biopolymers 105, 483 (2016)

work page 2016

[6] [6]

Krendel and M

M. Krendel and M. S. Mooseker, Physiology 20, 239 (2005)

work page 2005

[7] [7]

R. A. Muscat, J. Bath, and A. J. Turberﬁeld, Small 8, 3593 (2012)

work page 2012

[8] [8]

The special case of r→0 would realistically not allow motion in the positive direction, but permits simpliﬁca- tions in the state space of the system

Note that this deﬁnition implies that after the spider’s forward leg has cleaved the substrate site, it can take any number of half-steps back and forth from its original position before completing a whole step in either direction. The special case of r→0 would realistically not allow motion in the positive direction, but permits simpliﬁca- tions in the s...

work page