Stochastic Yield Catastrophe in Delay-Facilitated Self-Assembly

Erwin Frey; Richard Swiderski; Severin Angerpointner

REVIEW 2 major objections 4 minor 61 references

Stochastic Yield Catastrophe in Delay-Facilitated Self-Assembly

T0 review · 2 major / 4 minor · reviewed 2026-08-02 · deepseek-v4-flash

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

Pith's one-line read Delay-facilitated self-assembly in two coupled compartments suffers a stochastic yield catastrophe at low target numbers, caused by the random order of subunit and structure exchange; size-selective exchange restores yield.

desk verdict A genuine new result: delay-facilitated two-compartment assembly can hit a stochastic yield catastrophe at low target numbers when subunit and structure exchange are equally fast — but the biological window is narrower than the abstract suggests. read the letter →

arxiv 2607.13902 v2 pith:LKEOC7RA submitted 2026-07-15 physics.bio-ph

Richard Swiderski , Severin Angerpointner , Erwin Frey This is my paper · ORCID

classification physics.bio-ph

keywords yieldassemblyexchangenumbersstochastictargetstructuresdelay-facilitated

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

Cells often build large structures from many identical pieces, and sometimes this has to happen in tiny volumes with only enough material for a handful of structures, even one flagellum. A proposed way to make assembly work in unfavorable chemistry is to split the process between two connected compartments: a small 'fast' compartment where pieces bind quickly, and a large 'slow' compartment that slowly feeds pieces in. At large numbers, this delay trick works beautifully, and standard equations predict high yield.

This paper simulates the same two-compartment system when only a few structures can be made, for example enough material for just one finished ring of 30 subunits. The simulations show the trick fails: the final yield collapses. The reason is that pieces and half-built rings move between the compartments in random order. If the half-built ring drifts out of the fast compartment and two free pieces arrive before it returns, they can start a second, unwanted ring. With enough material for only one ring, that extra start is fatal. The effect is not a quirk of the particular chemistry: it happens even when each compartment on its own would assemble almost perfectly.

The paper then shows a fix: let small pieces move freely but restrict movement of larger chunks, for instance by letting exchange slow down with size. This restores most of the yield and does not change the average (mean-field) behavior. The same story appears for hexagonal building blocks and in a cell-membrane geometry where diffusion naturally makes bigger objects move more slowly.

Extended reading notes

Core claim

The paper's central claim: "Using stochastic simulations of a minimal two-compartment model, we show that delay-facilitated assembly is susceptible to a stochastic yield catastrophe at low target numbers: even when each compartment in isolation allows for high-yield assembly, slow exchange between them induces a substantial drop in the final yield." The asserted mechanism is that "the random order of comparably slow exchange events—subunit exchange versus structure exchange—determines whether the next reaction in the fast compartment is productive growth or excess nucleation." If correct, mean-field yield predictions for compartmentalized self-assembly can fail severely at low copy numbers, and the failure is controlled by the size dependence of exchange rates.

Load-bearing premise

The baseline catastrophe is demonstrated for size-independent exchange, D_n = D (Eq. 1e with D_n = D for all n), so partially built structures leave the fast compartment as readily as free subunits. The whole mechanism — a growing structure leaves, then two subunit entries cause excess nucleation — depends on this premise. The paper's own mitigation results (Sec. III C, Fig. 5) show that if structure exchange is even mildly slower (D_n = D_1/n) or absent (D_{n>1}=0), the yield drop essentially disappears. If real biological compartments are natively size-selective — as the authors suggest for pores — the baseline catastrophe would not occur in those systems, and the biological scope narrows to systems with non-selective exchange. The hybrid two-stage decomposition (Appendix D) is a second load-bearing approximation for the mechanistic attribution.

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

Colleague,

The paper is worth your time. The new result is real: the two-compartment delay-facilitated assembly strategy that works at mean-field level can fail badly at low target numbers, and the failure comes from the random order of comparably slow exchange events — subunit exchange versus structure exchange — in the second assembly stage, not from nucleation in either compartment alone. Prior work on well-mixed stochastic yield catastrophes [19] and on mean-field delay-facilitated assembly [26] did not cover this regime, so this is a genuine advance.

What's well done: the Gillespie simulations are carefully set up, with the mapping from mean-field to stochastic rates spelled out, bootstrap confidence intervals (width <0.005), and code deposited. The hybrid deterministic-first-stage/stochastic-second-stage simulation is a good diagnostic: it matches the full stochastic results closely, so the attribution to the second stage is credible. They also check robustness to structure size (S=120), hexagonal subunits, and a cytosol-membrane geometry, and they disclose the unrealistic boundary-diffusion caveat in Appendix G. The mitigation result — that slowing structure exchange restores yield without changing mean-field behavior — is clean and useful.

The soft spot is the one you flagged: the baseline catastrophe is for size-independent exchange D_n = D. If structure exchange is mildly slower (D_n = D_1/n) or absent, the yield drop mostly vanishes (Fig. 5). Since many real compartments are size-selective — pores, membrane binding — the raw catastrophe may be less common biologically than the abstract implies. The paper itself acknowledges this and positions the cytosol-membrane example as the realistic case. So the scope is narrower than 'slow exchange causes catastrophe'; more precisely, non-selective exchange causes it, and size-selective exchange is the fix. That is a scope limitation, not an internal inconsistency. The mechanism is argued qualitatively and supported by the hybrid simulation, not derived analytically; that is fine for a simulation study, though it limits the depth of the theory. The D→0 commensurability effect for high-yield compartments is a separate confound, but they clearly separate it.

Overall: central claim holds, the analysis is honest and reproducible. This deserves a serious referee; I'd accept it with a minor revision asking them to sharpen the abstract's generalization. I'd cite it in my own work on stochastic assembly robustness.

Referee Report

2 major / 4 minor

Summary. The paper extends a previously studied mean-field model of delay-facilitated self-assembly in two coupled compartments to the stochastic low-target-number regime. Using Gillespie simulations, it shows that at intermediate exchange rates the final yield drops dramatically for small target numbers N (N=1–10), even when each compartment in isolation supports high-yield assembly. The authors attribute this 'stochastic yield catastrophe' to a second-stage mechanism in which the random order of comparably slow subunit- and structure-exchange events can lead to excess nucleation. They further show that making structure exchange size-dependent (D_n = D_1/n) or suppressing it entirely (D_{n>1}=0) restores most of the yield without changing the mean-field behavior, and they report the same phenomenology for hexagonal subunits and for a cytosol-membrane geometry.

Significance. If the conclusions hold, this is an important contribution: it identifies a concrete failure mode of mean-field descriptions for compartmentalized self-assembly at biologically relevant low copy numbers, and it offers a design principle (size-selective exchange) to mitigate that failure. The paper's strengths include the direct stochastic simulations with bootstrap confidence intervals, a hybrid deterministic-stochastic decomposition that supports the mechanistic attribution, and reproducible code deposited on Zenodo. The central caveat is that the baseline catastrophe is established for size-independent exchange; the authors' own mitigation results show that even mild size selectivity removes the effect, which narrows the biological scope unless the claims are carefully qualified.

major comments (2)

[Abstract and Sec. IV B] The statements that 'delay-facilitated assembly is susceptible' to a stochastic yield catastrophe and that 'the same type of stochastic yield catastrophe emerges' for systems matching the model's basic assumptions overstate the scope. The catastrophe is demonstrated for D_n = D (Sec. III A, Fig. 2), while Fig. 5 shows that D_n = D_1/n or D_{n>1}=0 essentially eliminates it, and Appendix H shows these exchange modifications barely affect the mean-field behavior. Since the biological exchange mechanisms cited in Sec. IV B (pores, membrane binding) are often size-selective, the conditions for the catastrophe may be narrow. Please qualify the generalized claims, and ideally provide a quantitative map of the catastrophe's severity as a function of the size dependence of D_n.
[Appendix D and Fig. 4] The hybrid two-stage decomposition suppresses all first-stage stochasticity, including the initial subunit partition and nucleation in the fast compartment. For η = 10η*, first-stage stochasticity alone can cap the yield (Appendix E gives about 67% for N=1 with D_{n>1}=0). Consequently, the agreement between hybrid and fully stochastic simulations for η = 10η* in Fig. 4(b) does not by itself isolate the effect of the second stage: both curves could be low for different reasons. To solidify the mechanistic attribution, please quantify the separate first-stage and second-stage contributions to the yield loss, for example by comparing a hybrid with a stochastic first stage or by decomposing the yield gap between D_n = D and D_{n>1}=0.

minor comments (4)

[Appendix G / Sec. IV B] The cytosol-membrane simulation (Fig. 7) fixes the boundary diffusion at Dbar = 100, which the authors acknowledge is biologically unrealistic. Please add a sentence in Sec. IV B clarifying that this example is a proof-of-principle and may not be quantitatively representative for systems with slower boundary diffusion.
[Eq. (1e)] The sign convention in the exchange flux D_{n,α} = D_n(σ_{n,β} − σ_{n,α})/φ_α could be stated more explicitly: as written, D_{n,α} is the influx into compartment α from β. A one-sentence clarification would help readers.
[Abstract and Introduction] Typesetting artifacts such as 'O(10 4)' and 'O(10–104)' should be rendered as O(10^4) and O(10–10^4).
[Fig. 4] The legend distinguishes fully stochastic and hybrid simulations by 'small bullets' and 'large diamond markers'; in a black-and-white print these are easy to confuse. Consider adding different marker shapes/colors and a direct callout in the caption.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the yield catastrophe is a direct stochastic-simulation finding, tested by a hybrid variance decomposition and by mitigation controls; the mean-field inputs from the authors' prior work are reproduced in-paper and do not contain the low-target-number result.

full rationale

Walking the derivation chain: the central claim is the stochastic yield catastrophe at low target numbers in the two-compartment model (Sec. III A, Fig. 2). This is measured by direct Gillespie simulation (App. B), with rates fixed by a standard thermodynamic-limit mapping to the mean-field equations (Eq. B5); no parameter is fitted to reproduce the yield drop. The mechanism attribution—random ordering of subunit vs. structure exchange (Fig. 4a)—is supported by convergent, independent checks: (i) the deterministic-stochastic hybrid (App. D) reproduces the full stochastic result only because first-stage noise is non-essential, a nontrivial variance decomposition; (ii) the mitigation controls D_{n>1}=0 and D_n=D_1/n (Fig. 5) remove the proposed cause and suppress the catastrophe, a genuine causal test; (iii) the N=1 excess-nucleation argument and the ~67% cap (App. E) follow from stoichiometry and observed first-stage statistics, not from the yield data itself. The mean-field scaffolding (delay-facilitated regime; η=10η*, η*/10, τ=10^-6, φ_s=0.6) is taken from the same authors' prior work [19,26], but those results are published, and the present paper reproduces the needed mean-field curves in its own Fig. 1(e-f); their stated assumptions (mean-field, Eq. 1) do not include the low-N stochastic result, so the citations are real evidence rather than a self-citation chain. Stated limitations—the D→0 commensurability effect (Sec. III A, App. C), the biologically unrealistic boundary-diffusion rate in the cytosol-membrane illustration (App. G), and the fact that the baseline catastrophe uses size-independent D_n=D while mild size selectivity (D_n=D_1/n, D_{n>1}=0) removes it (Fig. 5)—are scope caveats, honestly disclosed, not circular steps. The hexagonal and bulk-boundary extensions (Figs. 6-7) are forward generalization tests; the bulk-boundary case imposes D_n=D_1/n, so its restored yield is consistent with Fig. 5 rather than an independent confirmation, but consistency of a model with its own earlier result is not circularity. No equation in the paper reduces to its own input by construction.

Assumptions & free parameters 5 free parameters · 9 assumptions · 0 invented entities

The central claim is a simulation result built from a Markovian model and its mean-field mapping. No new physical entities are postulated, and no parameters are fitted to produce the catastrophe. The hand-chosen inputs are representative, and robustness is checked across phi_s, S=120, hexagonal subunits, and a spatial geometry.

free parameters (5)

relative compartment reactivity tau = 10^-6
Chosen as an exemplary strongly slow compartment (tau << 1) where delay facilitation operates; not fitted to the catastrophe.
relative slow compartment volume phi_s = 0.6
Chosen above 0.5 where mean-field delay facilitation works; Figs. 2(c),(d) vary phi_s and show the catastrophe persists.
nucleation-to-growth ratio eta = 10 eta* and eta*/10
Representative low-yield and high-yield compartment conditions; the catastrophe appears for both, supporting generality.
target structure size S = 30 (and 120)
Main text uses S=30; Appendix I shows the same catastrophe for S=120.
boundary diffusion coefficient D_bar (cytosol-membrane model) = 100 (dimensionless)
Chosen large to avoid diffusion-limited boundary assembly; Appendix G discloses that this is biologically unrealistic.

assumptions (9)

standard math The stochastic process defined by the chemical master equation (B4) is exactly sampled by the Gillespie algorithm.
Standard stochastic chemical kinetics; used for all stochastic simulations in Sec. III.
domain assumption The deterministic mean-field equations (1) are the N -> infinity thermodynamic limit of the master equation with rate constants (B5).
Standard thermodynamic-limit assumption; the paper relies on it to compare stochastic and mean-field results.
domain assumption Linear ring assembly is irreversible and growth rate is independent of structure size.
Model assumption stated in Sec. II A; irreversibility makes excess nucleation permanently costly.
domain assumption Each compartment is well mixed, with symmetric exchange flux proportional to density differences (Eq. 1e).
Defines the minimal spatial model; no intra-compartment spatial gradients.
domain assumption The optimal nucleation-to-growth ratio eta* and its scaling from Ref. [19] apply to S=30 and S=120.
eta* is taken from prior work without re-derivation; Appendix I asserts the scaling.
domain assumption Mean-field delay-facilitated yield recovery in the intermediate-D regime from Ref. [26] holds as the baseline.
The paper's starting point, reproduced in Fig. 1(e)-(f).
ad hoc to paper The hybrid two-stage decomposition (Appendix D) is a valid approximation of the full process in the timescale-separated regime.
Introduces a first stage with D=0, tau=0 and a discretized second stage; the authors note it fails for D≲1 or tau≲1, but it is supported by agreement with full stochastic simulations.
domain assumption Hexagonal assembly reduces to the effective nucleation/growth equations (F1) with parameters from Ref. [51], including at low N.
The mapping is adopted from [51]; Fig. 6 notes finite-size deviations for N=100.
domain assumption In the cytosol-membrane geometry, size-dependent bulk diffusion D_n = D_1/n and a fast boundary diffusion isolate the fluctuation effect.
Appendix G sets D_bar=100, which is disclosed as biologically unrealistic but chosen to separate fluctuation effects from deterministic diffusion effects.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Stochastic Yield Catastrophe in Delay-Facilitated Self-Assembly." pith.science (2026). https://pith.science/paper/LKEOC7RA

@misc{pith2026260713902,
  author       = {Pith},
  title        = {Pith review of: Stochastic Yield Catastrophe in Delay-Facilitated Self-Assembly},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LKEOC7RA}},
  note         = {Machine review of arXiv:2607.13902}
}

read the original abstract

Self-assembly of supramolecular structures in cells and synthetic applications often proceeds under unfavorable biochemical conditions and at low copy numbers of final target structures, ranging from tens of bacterial microcompartments to a single bacterial flagellum per cell. Spatial organization through coupled reaction compartments of different reactivity (delay-facilitated assembly) can recover high yield in such environments at the mean-field level, but its robustness to stochastic fluctuations at low target numbers is unclear. Using stochastic simulations of a minimal two-compartment model, we show that delay-facilitated assembly is susceptible to a stochastic yield catastrophe at low target numbers: even when each compartment in isolation allows for high-yield assembly, slow exchange between them induces a substantial drop in the final yield. We trace the mechanism to a specific assembly stage, where the random order of rate-limiting exchange events of subunits and partially completed structures determines the ratio of productive growth to excess nucleation. Restricting the exchange of larger structures -- either by suppressing it entirely or letting exchange rates decrease with size -- restores most of the yield without altering the mean-field behavior. The same phenomenology appears for two-dimensional hexagonal subunits and in a cytosol-membrane geometry, where diffusion-limited exchange naturally implements the required size dependence. Our results show that equal success of assembly strategies at high target numbers does not imply their equal success at low target numbers, and that competing slow events occurring in random order are a common signature of stochastic yield catastrophes.

Figures

Figures reproduced from arXiv: 2607.13902 by the authors.

**Figure 2.** FIG. 2. Low target numbers cause a stochastic yield catastrophe. (a) Final yield for two coupled low-yield compartments [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗

**Figure 3.** FIG. 3. Mechanisms for robust assembly. (a) Snapshots [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (7 more)

**Figure 4.** Figure 4: FIG. 4. Mechanism leading to the delay-induced stochastic yield catastrophe in the two-compartment system. (a) Sketch of the [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: FIG. 5. Different exchange dynamics can alleviate the stochastic yield catastrophe. (a),(b) Final yield for two coupled low-yield [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]

**Figure 6.** Figure 6: FIG. 6. Assembly of two-dimensional subunits exhibit the same phenomenology. (a) Sketch of assembly reactions of hexagonal [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

**Figure 7.** Figure 7: FIG. 7. Diffusion-limited exchange recovers yield in a spa [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]

**Figure 8.** Figure 8: FIG. 8. Example state in the fast compartment after the first [PITH_FULL_IMAGE:figures/full_fig_p013_8.png]

**Figure 9.** Figure 9: FIG. 9. Delay-facilitated assembly unaffected by differ [PITH_FULL_IMAGE:figures/full_fig_p015_9.png]

**Figure 10.** Figure 10: FIG. 10. Delay-induced stochastic yield catastrophe for tar [PITH_FULL_IMAGE:figures/full_fig_p015_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

61 extracted references · 1 canonical work pages

[19]

M. F. Hagan, O. M. Elrad, and R. L. Jack, Mechanisms of kinetic trapping in self-assembly and phase transfor- mation, The Journal of Chemical Physics135, 104115 (2011). 16

2011
[26]

W. M. Jacobs, A. Reinhardt, and D. Frenkel, Rational design of self-assembly pathways for complex multicom- ponent structures, Proceedings of the National Academy of Sciences112, 6313 (2015)

2015
[1]

State space and elementary reactions The system is characterized at timetby the integer- valued vector of structure counts, N(t) = Nn,α(t) α∈{f,s}, n=1,...,S,(B1) whereN n,α is the number of structures of sizenin com- partmentα. Compartmentαhas volumeV α =ϕ αV; the total volumeV=SN/ρ 0 is fixed by the requirement that the system contain exactly enough mat...
[2]

Chemical master equation LetP(N, t) denote the probability that the system oc- cupies stateNat timet. From the state change propen- sities Eq.(B3) and the stoichiometric change vectors, the chemical master equation follows as, ∂tP(N, t) = X r h ar(N−∆ r)P(N−∆ r, t) −a r(N)P(N, t) i ,(B4) where the sum runs over all reactions r∈ {(nuc, α),(grow, n, α),(exc...
[3]

(B4) reduce to the deterministic equa- tions [Eq

Mapping to the mean-field parameters The microscopic rate constantsc nuc α ,c grow α , andc exch n,α are fixed by requiring that the mean of the dynamics generated by Eq. (B4) reduce to the deterministic equa- tions [Eq. (1)] in the thermodynamic limitN→ ∞at fixed densitiesσ n,α =N n,α/(Vαρ0). This matching re- quires cnuc α = 2τ α µ Vα = 2τ α η ν Vα ,(B5...
[4]

To determine a confidence inter- val around the reported final yield values for each data point we use a bootstrapping estimate with simple re- sampling [57]

Choice of sample size For stochastic simulations, we define the final yield as the average fraction of all subunits integrated into finished structures. To determine a confidence inter- val around the reported final yield values for each data point we use a bootstrapping estimate with simple re- sampling [57]. We dynamically choose the number of realizati...
[5]

Shajani, M

Z. Shajani, M. T. Sykes, and J. R. Williamson, Assembly of bacterial ribosomes, Annual Review of Biochemistry 80, 501 (2011)

2011
[6]

Baßler and E

J. Baßler and E. Hurt, Eukaryotic ribosome assembly, Annual Review of Biochemistry88, 1 (2018)

2018

Show all 61 references

[7]

Zlotnick and S

A. Zlotnick and S. Mukhopadhyay, Virus assembly, al- lostery and antivirals, Trends in Microbiology19, 14 (2011)

2011
[8]

J. D. Perlmutter and M. F. Hagan, Mechanisms of virus assembly, Annual Review of Physical Chemistry66, 1 (2015), 1407.3856

2015 arXiv
[9]

C. A. Kerfeld, C. Aussignargues, J. Zarzycki, F. Cai, and M. Sutter, Bacterial microcompartments, Nature Re- views Microbiology16, 277 (2018)

2018
[10]

F. F. V. Chevance and K. T. Hughes, Coordinating as- sembly of a bacterial macromolecular machine, Nature Reviews Microbiology6, 455 (2008)

2008
[11]

Nakamura and T

S. Nakamura and T. Minamino, Structure and dynamics of the bacterial flagellar motor complex, Biomolecules14, 1488 (2024)

2024
[12]

C. Sigl, E. M. Willner, W. Engelen, J. A. Kretzmann, K. Sachenbacher, A. Liedl, F. Kolbe, F. Wilsch, S. A. Aghvami, U. Protzer, M. F. Hagan, S. Fraden, and H. Di- etz, Programmable icosahedral shell system for virus trapping, Nature Materials20, 1281 (2021)

2021
[13]

Monferrer, F

A. Monferrer, F. Kohler, C. Sigl, M. Schachtner, D. Pe- terhoff, B. Asbach, R. Wagner, and H. Dietz, DNA origami traps for large viruses, Cell Reports Physical Sci- ence4, 101237 (2023)

2023
[14]

Khmelinskaia, A

A. Khmelinskaia, A. Wargacki, and N. P. King, Structure-based design of novel polyhedral protein nano- materials, Current Opinion in Microbiology61, 51 (2021)

2021
[15]

Jiang, Y

Q. Jiang, Y. Shang, Y. Xie, and B. Ding, DNA origami: from molecular folding art to drug delivery technology, Advanced Materials36, 2301035 (2024)

2024
[16]

Zlotnick, J

A. Zlotnick, J. M. Johnson, P. W. Wingfield, S. J. Stahl, and D. Endres, A theoretical model successfully identifies features of hepatitis B virus capsid assembly, Biochem- istry38, 14644 (1999)

1999
[17]

A. Y. Morozov, R. F. Bruinsma, and J. Rudnick, Assem- bly of viruses and the pseudo-law of mass action, The Journal of Chemical Physics131, 155101 (2009)

2009
[18]

M. F. Hagan and O. M. Elrad, Understanding the concen- tration dependence of viral capsid assembly kinetics—the origin of the lag time and identifying the critical nucleus size, Biophysical Journal98, 1065 (2010)

2010
[20]

Y. Ke, L. L. Ong, W. M. Shih, and P. Yin, Three- dimensional structures self-assembled from DNA bricks, Science338, 1177 (2012)

2012
[21]

B. Wei, M. Dai, and P. Yin, Complex shapes self- assembled from single-stranded DNA tiles, Nature485, 623 (2012)

2012
[22]

Reinhardt and D

A. Reinhardt and D. Frenkel, Numerical evidence for nu- cleated self-assembly of DNA brick structures, Physical Review Letters112, 238103 (2014)

2014
[23]

F. M. Gartner, I. R. Graf, P. Wilke, P. M. Geiger, and E. Frey, Stochastic yield catastrophes and robustness in self-assembly, eLife9, e51020 (2020)

2020
[24]

Rovigatti, J

L. Rovigatti, J. Russo, F. Romano, M. Matthies, L. Kroc, and P. ˇSulc, A simple solution to the problem of self- assembling cubic diamond crystals, Nanoscale14, 14268 (2022)

2022
[25]

D. E. P. Pinto, P. ˇSulc, F. Sciortino, and J. Russo, De- sign strategies for the self-assembly of polyhedral shells, Proceedings of the National Academy of Sciences120, e2219458120 (2023)

2023
[27]

M. F. Hagan and F. Mohajerani, Self-assembly coupled to liquid-liquid phase separation, PLOS Computational Biology19, e1010652 (2023)

2023
[28]

S. Laha, J. Bauermann, F. J¨ ulicher, T. C. T. Michaels, and C. A. Weber, Chemical reactions regulated by phase-separated condensates, Physical Review Research 6, 043092 (2024)

2024
[29]

Bartolucci, I

G. Bartolucci, I. S. Haugerud, T. C. Michaels, and C. A. Weber, The interplay between biomolecular assembly and phase separation, eLife13, 10.7554/elife.93003.3 (2026)

2026 doi
[30]

Angerpointner, R

S. Angerpointner, R. Swiderski, and E. Frey, Delay- facilitated self-assembly in compartmentalized systems, Proceedings of the National Academy of Sciences122, e2515123122 (2025), 2508.03383

2025
[31]

Phillips,Physical biology of the cell, 2nd ed

R. Phillips,Physical biology of the cell, 2nd ed. (Garland Science, London: New York, NY, 2013) pp. 39–49,121– 122

2013
[32]

Delbr¨ uck, The burst size distribution in the growth of bacterial viruses (bacteriophages), Journal of Bacteri- ology50, 131 (1945)

M. Delbr¨ uck, The burst size distribution in the growth of bacterial viruses (bacteriophages), Journal of Bacteri- ology50, 131 (1945)

1945
[33]

C. P. D. Brussaard, Viral control of phytoplankton populations—a review, Journal of Eukaryotic Microbi- ology51, 125 (2004)

2004
[34]

Parada, G

V. Parada, G. J. Herndl, and M. G. Weinbauer, Viral burst size of heterotrophic prokaryotes in aquatic sys- tems, Journal of the Marine Biological Association of the United Kingdom86, 613 (2006)

2006
[35]

H. Y. Chen, M. Di Mascio, A. S. Perelson, D. D. Ho, and L. Zhang, Determination of virus burst sizein vivousing a single-cycle SIV in rhesus macaques, Proceedings of the National Academy of Sciences104, 19079 (2007)

2007
[36]

N. C. Hill, J. W. Tay, S. Altus, D. M. Bortz, and J. C. Cameron, Life cycle of a cyanobacterial carboxysome, Science Advances6, eaba1269 (2020)

2020
[37]

Kojima, H

S. Kojima, H. Terashima, and M. Homma, Regulation of the single polar flagellar biogenesis, Biomolecules10, 533 (2020)

2020
[38]

Bange, G

G. Bange, G. Hochberg, K. Thormann, and A. Dornes, Where and how many: evolutionary diversification of a molecular switch regulating flagellar patterns, Journal of Bacteriology208, e00329 (2025)

2025
[39]

N. S. Tiwari and P. Van Der Schoot, Stochastic lag time in nucleated linear self-assembly, The Journal of Chemi- cal Physics144, 235101 (2016)

2016
[40]

T. C. T. Michaels, A. J. Dear, J. B. Kirkegaard, K. L. Saar, D. A. Weitz, and T. P. J. Knowles, Fluctuations in the kinetics of linear protein self-assembly, Physical Review Letters116, 258103 (2016)

2016
[41]

J. K. Davis and S. S. Sindi, Initial condition of stochastic self-assembly, Physical Review E93, 022109 (2016)

2016
[42]

Schaeffer, M

G. Schaeffer, M. J. Eleveld, J. Ottel´ e, P. C. Kroon, P. W. J. M. Frederix, S. Yang, and S. Otto, Stochastic emer- gence of two distinct self-replicators from a dynamic com- binatorial library, Journal of the American Chemical So- ciety144, 6291 (2022)

2022
[43]

L. B. Frechette, N. Sundararajan, F. Caballero, A. Tru- biano, and M. F. Hagan, Computer simulations show that liquid-liquid phase separation enhances self-assembly, ACS Nano19, 30275 (2025)

2025
[44]

D. T. Gillespie, A general method for numerically sim- ulating the stochastic time evolution of coupled chemi- cal reactions, Journal of Computational Physics22, 403 (1976)

1976
[45]

D. T. Gillespie, Stochastic simulation of chemical kinet- ics, Annual Review of Physical Chemistry58, 35 (2007)

2007
[46]

Chowdhury, S

C. Chowdhury, S. Chun, A. Pang, M. R. Sawaya, S. Sinha, T. O. Yeates, and T. A. Bobik, Selective molec- ular transport through the protein shell of a bacterial mi- crocompartment organelle, Proceedings of the National Academy of Sciences112, 2990 (2015)

2015
[47]

M. F. S. Lee, C. M. Jakobson, and D. Tullman-Ercek, Evidence for improved encapsulated pathway behavior in a bacterial microcompartment through shell protein engineering, ACS Synthetic Biology6, 1880 (2017)

2017
[48]

Chowdhury and T

C. Chowdhury and T. A. Bobik, Engineering the PduT shell protein to modify the permeability of the 1,2- propanediol microcompartment of salmonella, Microbi- ology165, 1355 (2019)

2019
[49]

Tasneem, T

N. Tasneem, T. N. Szyszka, E. N. Jenner, and Y. H. Lau, How pore architecture regulates the function of nanoscale protein compartments, ACS Nano16, 8540 (2022)

2022
[50]

T. H. Phan and E. N. G. Houben, Bacterial secretion chaperones: the mycobacterial type VII case, FEMS Mi- crobiology Letters365, fny197 (2018)

2018
[51]

Q. Xing, K. Shi, A. Portaliou, P. Rossi, A. Economou, and C. G. Kalodimos, Structures of chaperone-substrate complexes docked onto the export gate in a type III se- cretion system, Nature Communications9, 1773 (2018)

2018
[52]

Khmelinskaia, J

A. Khmelinskaia, J. M¨ ucksch, E. P. Petrov, H. G. Fran- quelim, and P. Schwille, Control of membrane binding and diffusion of cholesteryl-modified DNA origami nanos- tructures by DNA spacers, Langmuir34, 14921 (2018)

2018
[53]

Z. Hu, E. P. Gogol, and J. Lutkenhaus, Dynamic assem- bly of MinD on phospholipid vesicles regulated by ATP and MinE, Proceedings of the National Academy of Sci- ences99, 6761 (2002)

2002
[54]

C. F. Lang, O. Maxian, A. Anneken, and E. Munro, Oligomerization and positive feedback on membrane binding stabilize PAR-3 asymmetries in the C. elegans zygote, Current Biology36, 1509 (2026)

2026
[55]

F. M. Gartner and E. Frey, Design principles for fast and efficient self-assembly processes, Physical Review X14, 021004 (2024). 17

2024
[56]

B. J. H. M. Rosier, A. J. Markvoort, B. Gum ´ ı Aude- nis, J. A. L. Roodhuizen, A. Den Hamer, L. Brunsveld, and T. F. A. De Greef, Proximity-induced caspase-9 ac- tivation on a DNA origami-based synthetic apoptosome, Nature Catalysis3, 295 (2020)

2020
[57]

T. A. Leonard, M. Loose, and S. Martens, The membrane surface as a platform that organizes cellular and biochem- ical processes, Developmental Cell58, 1315 (2023)

2023
[58]

Rackauckas and Q

C. Rackauckas and Q. Nie, Differentialequations.jl–a per- formant and feature-rich ecosystem for solving differen- tial equations in Julia, Journal of Open Research Soft- ware5, 15 (2017)

2017
[59]

Gehring, D

J. Gehring, D. Widmann, D. Kleinschmidt, R. Finnegan, B. Kami´ nski, C. Bowers, H. Perez, M. Molignano, Milan Bouchet-Valat, P. K. Mogensen, T. Kelman, and N. Ig- natiadis, Juliangehring/Bootstrap.jl: Bootstrap v2.4.0, Zenodo (2023)

2023
[60]

Stochastic yield catastrophe in delay-facilitated self- assembly

R. Swiderski, S. Angerpointner, and E. Frey, Code for “Stochastic yield catastrophe in delay-facilitated self- assembly”’, Zenodo (2026)

2026
[61]

Efron, Nonparametric estimates of standard er- ror: The jackknife, the bootstrap and other methods, Biometrika68, 589 (1981)

B. Efron, Nonparametric estimates of standard er- ror: The jackknife, the bootstrap and other methods, Biometrika68, 589 (1981)

1981

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.

[1] [19]

M. F. Hagan, O. M. Elrad, and R. L. Jack, Mechanisms of kinetic trapping in self-assembly and phase transfor- mation, The Journal of Chemical Physics135, 104115 (2011). 16

2011

[2] [26]

W. M. Jacobs, A. Reinhardt, and D. Frenkel, Rational design of self-assembly pathways for complex multicom- ponent structures, Proceedings of the National Academy of Sciences112, 6313 (2015)

2015

[3] [1]

State space and elementary reactions The system is characterized at timetby the integer- valued vector of structure counts, N(t) = Nn,α(t) α∈{f,s}, n=1,...,S,(B1) whereN n,α is the number of structures of sizenin com- partmentα. Compartmentαhas volumeV α =ϕ αV; the total volumeV=SN/ρ 0 is fixed by the requirement that the system contain exactly enough mat...

[4] [2]

Chemical master equation LetP(N, t) denote the probability that the system oc- cupies stateNat timet. From the state change propen- sities Eq.(B3) and the stoichiometric change vectors, the chemical master equation follows as, ∂tP(N, t) = X r h ar(N−∆ r)P(N−∆ r, t) −a r(N)P(N, t) i ,(B4) where the sum runs over all reactions r∈ {(nuc, α),(grow, n, α),(exc...

[5] [3]

(B4) reduce to the deterministic equa- tions [Eq

Mapping to the mean-field parameters The microscopic rate constantsc nuc α ,c grow α , andc exch n,α are fixed by requiring that the mean of the dynamics generated by Eq. (B4) reduce to the deterministic equa- tions [Eq. (1)] in the thermodynamic limitN→ ∞at fixed densitiesσ n,α =N n,α/(Vαρ0). This matching re- quires cnuc α = 2τ α µ Vα = 2τ α η ν Vα ,(B5...

[6] [4]

To determine a confidence inter- val around the reported final yield values for each data point we use a bootstrapping estimate with simple re- sampling [57]

Choice of sample size For stochastic simulations, we define the final yield as the average fraction of all subunits integrated into finished structures. To determine a confidence inter- val around the reported final yield values for each data point we use a bootstrapping estimate with simple re- sampling [57]. We dynamically choose the number of realizati...

[7] [5]

Shajani, M

Z. Shajani, M. T. Sykes, and J. R. Williamson, Assembly of bacterial ribosomes, Annual Review of Biochemistry 80, 501 (2011)

2011

[8] [6]

Baßler and E

J. Baßler and E. Hurt, Eukaryotic ribosome assembly, Annual Review of Biochemistry88, 1 (2018)

2018