Pith. sign in

REVIEW 5 major objections 6 minor 21 references

Diffusive spreading of a polydisperse polymer solution in a channel

T0 review · 5 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read A single scaling law governs how long cut DNA takes to spread apart in a nanochannel, with the mean time growing as the total mass times the square of the required separation gap.

desk verdict A careful MC study of polydisperse single-file spreading with genuinely new observables and a real but fixable extrapolation gap; it deserves peer review with pressure to show the asymptotic analysis. read the letter →

arxiv 2506.08323 v1 pith:Y4HX6PLX submitted 2025-06-10 cond-mat.soft physics.bio-ph

Hanyang. Wang , Gary W Slater This is my paper
classification cond-mat.softphysics.bio-ph
keywords single-filediffusionspreadingtimepolydispersepolymerDNAmappingnanochannelfirst-passageMonteCarlosimulationRousemodel
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper asks how long it takes for a long DNA molecule, cut into fragments inside a narrow channel, to spread far enough that the fragments can be imaged separately. Treating the process as single-file diffusion, the authors use lattice Monte Carlo simulations to show that the mean spreading time scales as $\tau_f \sim M[\Delta s]^2$, where $M$ is the total molecular mass and $\Delta s$ is the minimum required increase in the system's span. They also find that the rescaled spreading time $t = \tau_f / \langle \tau_f \rangle$ follows a universal inverse-Gaussian-like distribution whose width shrinks as $1/\sqrt{N}$. Crucially, the randomness of Brownian motion contributes as much to the spread of spreading times as the specific sequence of fragment sizes, so individual spreading times cannot be reliably predicted from sequence details alone.

What carries the argument

The key object is the length scale $\Delta s = (N-1)(\tilde X-1)$, the minimum increase of the total span $S(t)$ required for all $N-1$ gaps to reach the detection threshold $\tilde X$ simultaneously. The argument is carried by a lattice Monte Carlo model of single-file diffusion with Rouse-like mobilities $D(m_i) = D_1/m_i$, where the first-passage spreading time $\tau_f$ is the first time at which all gaps $X_i$ exceed $\tilde X$. The scaling collapses of $\langle\tau_f\rangle$, the final span increase $\Delta S_f$, and the center-of-mass displacement onto combinations of $\Delta s$, $M$, and $N$, together with the universal collapse of the rescaled distribution, provide the mechanism that links microscopic parameters to observable waiting times.

What would settle it

In a nanochannel experiment with a DNA molecule cut into $N$ fragments of known sizes, repeatedly measure the time until all gaps first exceed a set threshold $\tilde X$. If the mean does not grow as $M(N-1)^3(\tilde X-1)^2$ and the rescaled distribution's standard deviation does not fall as $1/\sqrt{N}$, the central scaling claim is refuted.

Watch

Extended reading notes

Core claim

The central claim is that the mean spreading time for polydisperse single-file fragments obeys $\langle \tau_f \rangle \sim M[\Delta s]^2 \sim m(N-1)^3(\tilde X-1)^2$, with $\Delta s = (N-1)(\tilde X-1)$ the minimum span increase needed for all gaps to exceed the detection threshold $\tilde X$. This asymptotic scaling holds for both monodisperse and random fragment-size distributions, and it shows that the spreading time is the time an uncut molecule of mass $M$ would need to diffuse over a distance $\Delta s$. The paper further establishes that the distribution of rescaled spreading times $t = \tau_f/\langle\tau_f\rangle$ is universal across system parameters, is well described by a first-passage-inspired inverse-Gaussian form, and has a standard deviation that falls as $1/\sqrt{N}$. Because the sequence-dependent shift in mean spreading time is comparable to the Brownian spread, the paper concludes that the stochastic nature of diffusion is as significant as the molecular size distribution in determining any individual spreading time.

Load-bearing premise

The scaling laws assume that fragment diffusion follows the Rouse rule $D(m)=D_1/m$ with rigid fragments and no internal length fluctuations; if real DNA fragments in a channel diffuse with a different size exponent, the predicted exponents in the spreading-time law would change.

Editorial extensions

If this is right

  • In a nanochannel DNA-mapping experiment, the wait for fragments to become resolvable grows as total DNA length times the square of the required gap, so doubling the number of cuts can increase the wait by roughly an order of magnitude.
  • The universal rescaled distribution means the relative uncertainty in any single spreading time is set only by the number of fragments, about $1/\sqrt{N}$, independent of the fragment sizes and the channel resolution.
  • The $Z$ parameter gives a linear ranking of expected spreading times: placing large fragments near the channel ends slows spreading, while centralizing large fragments speeds it up.
  • Since Brownian variability is as large as sequence effects, the mean spreading time is a useful guide for ensemble behavior but not a reliable predictor for a specific cut molecule.
  • The asymptotic scaling $\langle\tau_f\rangle \sim M[\Delta s]^2$ provides a simple formula for estimating the required imaging time from known experimental parameters.

Reading between the lines

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

  • The universality of the rescaled distribution suggests that the spreading time behaves like the first-passage time of an effective Brownian particle moving over a distance set by $\Delta s$, which may allow the same collapse to apply to other single-file systems with size-dependent mobilities.
  • If real DNA fragment diffusion inside a nanochannel follows a size exponent different from Rouse ($D \sim 1/m$), the predicted exponents in the spreading-time scaling law should shift accordingly, offering a way to measure in-channel diffusion exponents from waiting-time statistics.
  • The $Z$ parameter could serve as a design criterion: cutting schemes that keep the largest fragments near the center of the channel would minimize $Z$ and thereby shorten the mapping protocol in practice.
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

5 major / 6 minor

Summary. The paper studies the single-file diffusion of polydisperse polymer fragments in a narrow channel, motivated by restriction-enzyme DNA mapping in nanochannels. Using lattice Monte Carlo simulations, the authors define a first-passage spreading time tau_f as the time when all inter-fragment gaps first exceed a threshold Xtilde. They report scaling laws for the mean spreading time, the span increase, and the center-of-mass displacement, with proposed asymptotic forms <tau_f> ~ m N^3 Xtilde^2 ~ M [Delta s]^2 and Delta S_f ~ N Delta s, where Delta s=(N-1)(Xtilde-1). They introduce a sequence-randomness parameter Z that correlates with tau_f, and they show that the rescaled spreading time t=tau_f/<tau_f> follows a universal inverse-Gaussian-like distribution whose standard deviation decreases as 1/sqrt(N). The overall conclusion is that Brownian variability is comparable to sequence effects in determining individual spreading times.

Significance. If the proposed asymptotic scaling laws hold, the paper identifies a single fundamental length scale Delta s that controls the spreading dynamics of polydisperse single-file systems, which would be of practical value for designing nanochannel DNA mapping experiments. The universal distribution of rescaled spreading times, with a 1/sqrt(N) decay of its width, is a potentially useful quantitative prediction. The paper also makes a valuable conceptual contribution by demonstrating that stochastic diffusion noise is at least as important as the specific molecular sequence for predicting individual spreading times, and the Z parameter provides a simple heuristic for ordering sequences by expected spreading time. The simulation study is extensive, with large ensembles and reported statistical errors, and the paper is clearly written. However, the central asymptotic claims rest on extrapolations that are stated as 'data not shown', and the finite-range effective exponents still differ substantially from the claimed asymptotic values; this is the main weakness.

major comments (5)
  1. [§III.A, Eqs. (4)–(5)] The central asymptotic scaling law <tau_f> ~ m N^3 Xtilde^2 is asserted on the basis of 'Extrapolating our results (data not shown)'. The finite-range fits quoted in Eq. (4) give eta_tau=2.48(2) and chi_tau=1.73(1), which are 20–25% away from the claimed asymptotic values 3 and 2. The manuscript must display the extrapolation curves, specify the range of N and Xtilde used, the assumed functional form for the approach to the asymptote, and the resulting uncertainties. The same 'data not shown' extrapolation supports Eqs. (8), (11), and (14)–(16); without this evidence, the headline scaling law is not supported by the data actually presented.
  2. [§IV (Conclusion)] The 'derivation' of tau_f ~ M [Delta s]^2 is a dimensional consistency argument that presupposes the scaling relation it is used to justify: it assumes <Delta x_CM^2(tau_f)> ~ [Delta s]^2 and <tau_f> ~ <Delta x_CM^2>/D(M). The CM displacement scaling is itself an extrapolated simulation result. Please either provide an independent derivation (e.g., from a first-passage or single-file diffusion theory) or explicitly label this reasoning as a heuristic consistency check rather than a proof.
  3. [§III.A, Eq. (12) and later usage] The prefactors 1/2 and 1/5 in Eq. (12) are fixed from a single simulation point (N=11, m=10, Xtilde=15) and later used as 'predictions' in Figs. 5 and 6. This is circular when the same data are used to calibrate and then to validate. The text should state clearly that these are empirical estimates with no quantified uncertainty, and quantitative comparisons with simulation data should be provided rather than qualitative statements like 'an excellent estimate overall'.
  4. [§II, initial conditions] The model sets all initial gaps equal to one lattice cell, X_i(0)=1. In an experimental restriction-enzyme cut, the gaps after cutting will not be uniform, and first-passage times can be sensitive to the initial configuration. The paper does not test how the results change with a distribution of initial gaps; at minimum, this assumption should be explicitly acknowledged as an idealization whose effect on the scaling laws and on the spread of tau_f is unknown.
  5. [§II, diffusion model and §III.A scaling] The predicted exponents rely on the Rouse scaling D(m)=D_1/m and on neglecting axial length fluctuations. If real DNA fragments in a nanochannel follow a different size-diffusivity exponent, or if internal polymer modes matter, the scaling laws in Eqs. (5), (11), and (15) would change. The paper should either justify the Rouse assumption for the relevant parameter range (citing experiment or theory) or discuss how the exponents would shift under plausible alternative models.
minor comments (6)
  1. [Abstract] The phrase 'its variance decreases linearly with the number of fragments' is incorrect: Fig. 9(d) shows the standard deviation decreasing as 1/sqrt(N), so the variance decreases as 1/N, i.e., inversely with N, not linearly with N.
  2. [§III.E] Typo: 'minium fragment size' should be 'minimum fragment size'.
  3. [§III.B, around Eq. (13)] The text states 'with eta_tau approximately 2.75 (not shown)' for the bimodal data, whereas Eq. (4) gives eta_tau=2.48(2) and the asymptotic extrapolation is 2.93(5). Please clarify which exponent is used in Eq. (13) and why it differs from the monodisperse value.
  4. [§III.C, Eq. (17)] The definition of Z is given only for N odd. Since experimental systems may have even N, a definition for even N should be provided, or the text should state how the formula is generalized.
  5. [FIG. 2 inset] The inset plots <Delta x_CM^2(tau_f)>/<Delta S_f>^2 versus N/(N-1)^3, but the text does not explain why the abscissa is chosen in this form or what the expected behavior is.
  6. [References] Reference [15] is listed as 'B. H. C., Random Walks in Biology'; the author is H. C. Berg, and the reference should be given in standard form.

Circularity Check

2 steps flagged · score 3.0 of 10

Central scaling derivation is empirical and not circular, but two fitted parameters are loosely called predictions.

  1. fitted input called prediction [Section III E, discussion of Eq. (18) and Fig. 9]
    "Given that the standard deviation of our data is 0.466(7), this predicts t0 = 2.30(7) and a maximum at t = 0.726(7)."

    Equation 18 defines P(t) with variance 1/(2t0). The observed standard deviation 0.466(7) therefore fixes t0 = 1/(2×0.466²) ≈ 2.30; the variance is an input, not a prediction. Only the mode location 0.726(7) follows from the assumed functional form. The later claim that Eq. 18 provides an excellent description is a genuine shape comparison, but agreement of the variance is automatic by construction.

  2. fitted input called prediction [Section III A, Eq. (12); later invoked in Section III B]
    "It is possible to estimate the missing asymptotic coefficients in Eqs. 5 and 11 using the fact that ⟨τ f ⟩ ≈ 10^6 and ∆S f ≈ 300 in Fig. 3. Since M = mN = 10×11 = 110 while ∆s = (N−1)(X̃−1) = 14×10 = 140, we obtain ⟨τ f ⟩ ≈ 1/2 M[∆s]^2, ∆S f ≈ 1/5 N∆s. ... Equation 12 predicts ⟨τ f ⟩ ≈ 0.83 10^5, close to the lowest plateau."

    The constants 1/2 and 1/5 are obtained by imposing the formula on a single simulation point (⟨τf⟩≈10^6, ΔSf≈300 from Fig. 3). When the text later says 'Equation 12 predicts ⟨τf⟩≈0.83×10^5' for a bimodal system, this is an evaluation of the same calibrated relation, not an independent prediction. It tests internal consistency of the assumed scaling form, but the 'prediction' is statistically tied to the calibration point.

full rationale

The central scaling laws (Eqs. 4, 5, 14, 15) are obtained by direct fits to independent Monte Carlo data; the asymptotic exponents are extrapolations ('data not shown') rather than derivations, so their evidential status is weaker than claimed, but extrapolation is not circular. The Z-parameter is defined without reference to τ_f and its correlation with τ_f is tested against simulation. Eq. 18 is an empirical ansatz borrowed from single-particle first-passage theory; the shape comparison is meaningful, but t0 is fixed by the observed variance rather than predicted. Eq. 12 calibrates coefficients with a single simulation point; later uses of Eq. 12 are evaluations of that calibrated formula. No load-bearing self-citation: refs [12–14] are method/physics citations, and [14] is a co-authored algorithm reference but not used to justify the scaling claims. Overall the derivation chain is not circular; the two flagged items are wording/calibration issues, giving a score of 3.

Assumptions & free parameters 13 free parameters · 5 assumptions · 1 invented entities

The scaling laws are empirical: the exponents are fitted to Monte Carlo data, not derived from first principles. The model imports Rouse-type rigid-blob diffusion assumptions, and the Z parameter is an ad hoc descriptor with limited predictive power. The central simulation outputs are self-contained, but the asymptotic extrapolations and universality claims are fit-extended.

free parameters (13)
  • eta_tau (monodisperse) = 2.48(2) finite range; extrapolated to ~2.93(5)
    Power-law exponent for N in Eq. (4), obtained by fitting Monte Carlo data.
  • chi_tau (monodisperse) = 1.73(1) finite range; extrapolated to ~1.99(4)
    Power-law exponent for tilde X in Eq. (4), fitted to Monte Carlo data.
  • eta_tau_R (random) = 2.60(1)
    Exponent for N in Eq. (14), fitted to random-sequence Monte Carlo data.
  • chi_tau_R (random) = 1.70(5)
    Exponent for tilde X in Eq. (14), fitted to random-sequence Monte Carlo data.
  • eta_s (span, monodisperse) = 1.54(1)
    Exponent in Eq. (10) for the increase of the span.
  • chi_s (span, monodisperse) = 0.975(6)
    Exponent in Eq. (10) for the increase of the span.
  • eta_s_R (span, random) = 1.52(3)
    Exponent in Eq. (16) for random sequences.
  • chi_s_R (span, random) = 0.88(1)
    Exponent in Eq. (16) for random sequences.
  • eta_x (CM displacement) = 0.6(1)
    Exponent in Eq. (7) for center-of-mass displacement.
  • chi_x (CM displacement) = 1.8(1)
    Exponent in Eq. (7) for center-of-mass displacement.
  • t0 (distribution shape) = 2.30(7)
    Parameter in Eq. (18), determined by matching the observed variance of the rescaled distribution.
  • Slope of tau_f versus Z = 0.47(2) to 0.49(3) times 10^6
    Linear fit to binned data in Figs. 6(b) and 7; the correlation is weak relative to the spread.
  • Approximate prefactors in Eq. (12) = 1/2 for tau_f and 1/5 for Delta S_f
    Estimated from a single simulation point in Fig. 3 and then used as general approximations.
assumptions (5)
  • domain assumption Each fragment's diffusion coefficient scales as D(m_i) = D_1 / m_i, the Rouse model.
    Section II states this relation directly. It is imported from polymer physics and is central to the polydispersity effects.
  • domain assumption Fragments are rigid objects on a lattice and axial length fluctuations are neglected.
    Section II: 'the fluctuations in the axial length of the molecules are neglected'. Internal polymer dynamics are ignored.
  • ad hoc to paper Initial gaps are exactly one lattice cell, X_i(0) = 1 for all i.
    Section II sets this idealized initial condition. Real cutting may produce non-uniform initial gaps.
  • ad hoc to paper Minimum detectable gap and minimum visible fragment size are both equal to tilde X.
    Section III E says the minimum final gap and minimum fragment size are both denoted tilde X; only one alternative case is tested.
  • domain assumption Hydrodynamic interactions between blobs are screened and single-file ordering prevents fragments from passing.
    Section II cites refs. [12,13] for this standard nanochannel blob-model assumption.
invented entities (1)
  • Z, the molecular sequence randomness or weight-torque parameter
    purpose: Quantifies how much mass sits near the ends versus the center of the fragment sequence; used to correlate sequence order with spreading time.
    Defined ad hoc in Eq. (17) for odd N. The paper itself reports that the correlation is weak relative to Brownian fluctuations and that Z cannot reliably predict individual spreading times, so no strong external falsifiable handle is established.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Diffusive spreading of a polydisperse polymer solution in a channel." pith.science (2026). https://pith.science/paper/Y4HX6PLX

@misc{pith2026250608323,
  author       = {Pith},
  title        = {Pith review of: Diffusive spreading of a polydisperse polymer solution in a channel},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Y4HX6PLX}},
  note         = {Machine review of arXiv:2506.08323}
}
abstract

Long DNA molecules can be mapped by cutting them with restriction enzymes inside a narrow channel. Once cut, the individual fragments thus produced move away from each other due to diffusion and entropic effects. We investigate how long it takes for these fragments to travel distances large enough for an experimental device to distinguish them and (possibly) estimate their size. In essence, this is a single-file diffusion process in which molecules of different sizes and hence different diffusion coefficients spread out from an initially dense configuration. We use Monte Carlo methods to investigate this class of problems and define the time taken to reach the required final state as a first-passage \textit{spreading time}. Our results demonstrate that the stochastic nature of the diffusion process is as significant as the specifics of the molecular size distribution in determining the spreading time. We examine the relationship between the spreading time and the final space occupied by the fragments as a function of the experimental parameters and determine the fundamental length scale governing this process. We introduce a molecular sequence randomness parameter, $Z$, which is linearly correlated with the final spreading time. Finally, we show that the distribution function of spreading times follows a well-known form for first-passage time problems, and that its variance decreases linearly with the number of fragments.

Figures

Figures reproduced from arXiv: 2506.08323 by the authors.

Figure 1
Figure 1. FIG. 1. Diagram of the molecular model and its mapping onto the [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Scaled mean spreading time [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. The data points [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (5 more)
Figure 5
Figure 5. Figure 5: (a) shows ⟨τf⟩ vs. the mass ratio µ for a sys￾tem where N1 = 3, N2 = 2, M = 648 and P = 1 are kept FIG. 5. Mean spreading time ⟨τf ⟩ vs mass ratio µ = m2/m1. The systems are described symbolically as series of 1 and 2. The total molecular weight is fixed at M = 648, wh…
Figure 6
Figure 6. Figure 6: FIG. 6. (a) Spreading time [PITH_FULL_IMAGE:figures/full_fig_p005_6.png]
Figure 7
Figure 7. Figure 7: (a) shows ⟨τf⟩ vs Z data for four initial sequences. The linear increase observed in [PITH_FULL_IMAGE:figures/full_fig_p006_7.png]
Figure 8
Figure 8. Figure 8: (c) shows that the distributions do not overlap for the low-τf outliers, all of which have large objects near their midpoints. The shortest times are found for the smallest Z value (in green): the central molecule (size 362) is over 50% of the total weight. The largest…
Figure 9
Figure 9. Figure 9: FIG. 9. (a) Distribution function of rescaled spreading times [PITH_FULL_IMAGE:figures/full_fig_p007_9.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

21 extracted references · 21 canonical work pages

  1. [1]

    Karolin, M

    F. Karolin, M. Vilhelm, K. Sriram, D. D. Kevin, and W. Fredrik, Quarterly Reviews of Biophysics55(e12), 1 (2022)

    work page 2022

  2. [2]

    weight torque

    What is the effect of the mass asymmetryµ? Figure 5 (a) shows⟨τ f ⟩vs.the mass ratioµfor a sys- tem whereN 1 =3,N 2 =2,M=648 andP=1 are kept FIG. 5. Mean spreading time⟨τ f ⟩vsmass ratioµ=m 2/m1. The systems are described symbolically as series of1and 2. The total molecular weight is fixed atM=648, while ˜X=5 andN 1 =3. (a)N 2 =2 andP=1; (b)N 2 =2Pwith, f...

    work page 2000

  3. [3]

    In all three cases, the lowest data set is forN1 =0, the dashed horizontal lines correspond toN 2 =0, while the dotted lines present the[N 1 =1;m 1 →∞]limit

    The next three panels show the mean spreading time⟨τ f ⟩vsthe numberN 2 of type 2 molecules of sizem 2 =10 for three different values ofm 1; the final minimum spacing is ˜X=5 and the ensemble size is 12 000. In all three cases, the lowest data set is forN1 =0, the dashed horizontal lines correspond toN 2 =0, while the dotted lines present the[N 1 =1;m 1 →...

    work page

  4. [4]

    D. C. Schwartz, X. Li, L. I. Hernandez, S. P. Ramnarain, E. J. Huff, and Y . K. Wang, Science262, 110 (1993)

    work page 1993

  5. [5]

    D. C. Schwartz, X. Li, L. I. Hernandez, S. P. Ramnarain, E. J. Huff, and Y . K. Wang, Nature genetics9, 432 (1995)

    work page 1995

  6. [6]

    Reisner, N

    W. Reisner, N. B. Larsen, A. Silahtaroglu, A. Kristensen, N. Tommerup, J. O. Tegenfeldt, and H. Flyvbjerg, Proc Natl Acad Sci107, 13294 (2010)

    work page 2010

  7. [7]

    T. Erik, G. Gaurav, J. Anna, W. Fredrik, and A. Tobias, PloS one16, e0260489 (2021)

    work page 2021

  8. [8]

    Gruszka, J

    D. Gruszka, J. Jeffet, S. Margalit, Y . Michaeli, and Y . Eben- stein, Essays in Biochemistry65, 51 (2021)

    work page 2021

Show all 21 references
  1. [9]

    S. M. Friedrich, H. C. Zec, and T. Wang, LAB ON A CHIP16, 9 790 (2016)

    2016

  2. [10]

    Marzia, V

    I. Marzia, V . N. Chandra, B. Ivy, and M. Amit, Analytical Chemistry97(16), 8641 (2025)

    2025

  3. [11]

    X. Meng, K. Benson, K. Chada, E. J. Huff, and D. C. Schwartz, Proc Natl Acad Sci102, 10012 (2005)

    2005

  4. [12]

    Uppuluri, T

    L. Uppuluri, T. Jadhav, Y . Wang, and M. Xiao, Analytical Chemistry93(28), 9808 (2021)

    2021

  5. [13]

    McCaffrey, J

    J. McCaffrey, J. Sibert, B. Zhang, Y . Zhang, W. Hu, H. Rieth- man, and M. Xiao, Nucleic Acids Research44(2), e11 (2015)

    2015

  6. [14]

    Brochard and P

    F. Brochard and P. G. de Gennes, J. Chem. Phys.67, 52 (1977)

    1977

  7. [15]

    Daoud and P

    M. Daoud and P. G. D. Gennes, Journal de Physique38, 85 (1977)

    1977

  8. [16]

    M. V . Chubynsky and G. W. Slater, Physical review. E85, 016709 (2012)

    2012

  9. [17]

    B. H. C.,Random Walks in Biology.(Princeton University Press, Princeton, N.J., 1993)

    1993

  10. [18]

    Y . Koki, O. Nobutoshi, and T. Yo, Anal. Chem.93, 332 (2021)

    2021

  11. [19]

    S. Zhou, F. Wei, J. Nguyen, M. Bechner, K. Potamousis, S. Goldstein, L. Pape, M. R. Mehan, C. Churas, S. Pasternak, D. K. Forrest, R. Wise, D. Ware, R. A. Wing, M. S. Waterman, M. Livny, and D. C. Schwartz, PLoS genetics5(11), e1000711 (2009)

    2009

  12. [20]

    Teague, M

    B. Teague, M. S. Waterman, S. Goldstein, K. Potamousis, S. Zhou, S. Reslewic, D. Sarkar, A. Valouev, C. Churas, J. M. Kidd, S. Kohn, R. Runnheim, C. Lamers, D. Forrest, M. A. Newton, E. E. Eichler, M. Kent-First, U. Surti, M. Livny, and D. C. Schwartz, Proc Natl Acad Sci107, 1...

    2010

  13. [21]

    Ebert, P

    P. Ebert, P. Audano, B. Zhu, Q. Rodriguez-Martin, D. Porubsky, and M. e. a. Bonder, Science372, eabf7117 (2021)

    2021

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.