Skip to main content
Log in

Approximate majority analyses using tri-molecular chemical reaction networks

  • Published:
Natural Computing Aims and scope Submit manuscript

Abstract

Approximate Majority is a well-studied problem in the context of chemical reaction networks (CRNs) and their close relatives, population protocols: Given a mixture of two types of species with an initial gap between their counts, a CRN computation must reach consensus on the majority species. Angluin, Aspnes, and Eisenstat proposed a simple population protocol for Approximate Majority and proved correctness and \(O(\log n)\) time efficiency with high probability, given an initial gap of size \(\omega (\sqrt{n}\log n)\) when the total molecular count in the mixture is n. Motivated by their intriguing but complex proof, we provide a new analysis of several CRNs for Approximate Majority, starting with a very simple tri-molecular protocol with just two reactions and two species. We obtain simple analyses of three bi-molecular protocols, including that of Angluin et al., by showing how they emulate the tri-molecular protocol. Our results improve on those of Angluin et al. in that they hold even with an initial gap of \(\varOmega (\sqrt{n \log n})\). We prove that our tri-molecular CRN is robust even when there is some uncertainty in the reaction rates, when some molecules are Byzantine (i.e., adversarial), or when activation of molecules is triggered by epidemic. We also analyse a natural variant of our tri-molecular protocol for the more general problem of multi-valued consensus. Our analysis approach, which leverages the simplicity of a tri-molecular CRN to ultimately reason about these many variants, may be useful in analysing other CRNs too.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

Notes

  1. Here is the calculation for the probability conversion.

    $$\begin{aligned} \rho _{r}(c)&= k'_{r} . \left[ \prod _{i=1}^m (x_i !/(x_i - s_i)!)\right]/v^{o-1} \\&= k'_{r} . \left[ \prod _{i=1}^m s_i!\right] . \left[ \prod _{i=1}^m \left( {\begin{array}{c}x_i\\ s_i\end{array}}\right) \right]/v^{o-1} \\&= \left[ \left( {\begin{array}{c}n\\ o\end{array}}\right) /v^{o-1} \right] k'_{r} \left[ \prod _{i=1}^m s_i!\right] . \left[ \prod _{i=1}^m \left( {\begin{array}{c}x_i\\ s_i\end{array}}\right) \right]/\left( {\begin{array}{c}n\\ o\end{array}}\right) \\&= \left[ \left( {\begin{array}{c}n\\ o\end{array}}\right) /v^{o-1} \right] k_{r} \left[ \prod _{i=1}^m \left( {\begin{array}{c}x_i\\ s_i\end{array}}\right) \right]/\left( {\begin{array}{c}n\\ o\end{array}}\right) , \end{aligned}$$

    where

    $$\begin{aligned} k_{r} = k'_{r} \left[ \prod _{i=1}^m s_i!\right] . \end{aligned}$$
    (1)

    We can interpret the last of these expressions for \(\rho _{r}(c)\) as the product of three terms. The first term, namely \(\left( {\begin{array}{c}n\\ o\end{array}}\right) /v^{o-1}\), corresponds to the (normalized) average rate of an interaction of order \(o\). The last term, namely \([\prod _{i=1}^m \left( {\begin{array}{c}x_i\\ s_i\end{array}}\right) ]/\left( {\begin{array}{c}n\\ o\end{array}}\right) \), is the probability that the reaction of order \(o\) has exactly the reactants of \(r\). The middle term \(k_{r}\) depends on the \(s_i\)’s, but could also model situations where different types of interactions have different rates, e.g., if some molecular species are larger than others. Normalizing the \(k_{r}\)’s by \(\sum k_{r}\) yields rate constants for our model.

  2. Here, we tacitly assume that \(n=(2k)^2\), for some integer k. Extending the argument to general n, though notationally cumbersome, is straightforward.

  3. A similar, though more involved argument, can be used to show the same result with relaxed X-consensus defined as an X-population of size \(n - (1+\epsilon )z\), for any \(\epsilon > 0\).

References

  • Alistarh D, Aspnes J, Eisenstat D, Gelashvili R, Rivest RL (2017) Time-space trade-offs in population protocols. In: Proceedings of the twenty-eighth annual ACM-SIAM symposium on discrete algorithms. pp 2560–2579

  • Angluin D, Aspnes J, Diamadi Z, Fischer MJ, Peralta R (2006a) Computation in networks of passively mobile finite-state sensors. Distrib Comput 18(4):235–253

    Article  Google Scholar 

  • Angluin D, Aspnes J, Eisenstat D (2006b) Fast computation by population protocols with a leader. In: Dolev S (ed) Distributed computing (DISC), vol 4167. Lecture notes in computer science. Springer, Berlin, pp 61–75

    Chapter  Google Scholar 

  • Angluin D, Aspnes J, Eisenstat D (2008) A simple population protocol for fast robust approximate majority. Distrib Comput 21(2):87–102

    Article  Google Scholar 

  • Becchetti L, Clementi A, Natale E, Pasquale F, Silvestri R, Trevisan L (2016) Simple dynamics for plurality consensus. Distrib Comput 30:1–14

    MathSciNet  MATH  Google Scholar 

  • Becchetti L, Clementi AEF, Natale E, Pasquale F, Trevisan L (2016) Stabilizing consensus with many opinions. In: Proceedings of the twenty-seventh annual ACM-SIAM symposium on discrete algorithms. pp 620–635

  • Cardelli L, Csikász-Nagy A (2012) The cell cycle switch computes approximate majority. Nat Sci Rep 2:656

    Article  Google Scholar 

  • Cardelli L, Kwiatkowska M, Laurenti L (2016) Programming discrete distributions with chemical reaction networks. In: Rondelez Y, Woods D (eds) DNA computing and molecular programming, vol 9818. Lecture notes in computer science. Springer, Cham, pp 35–51

    Chapter  Google Scholar 

  • Chen Y-J, Dalchau N, Srinivas N, Phillips A, Cardelli L, Soloveichik D, Seelig G (2013) Programmable chemical controllers made from DNA. Nat Nanotechnol 8(10):755–762

    Article  Google Scholar 

  • Chernoff H (1952) A measure of asymptotic efficiency for tests of a hypothesis based on the sum of observations. Ann Math Stat 23:493–507

    Article  MathSciNet  Google Scholar 

  • Condon A, Hajiaghayi M, Kirkpatrick D, Manuch J (2017) Simplifying analyses of chemical reaction networks for approximate majority. In: 23rd international conference on DNA computing and molecular programming (Lecture notes in computer science), vol 10467. Springer-Verlag, pp 189–209

  • Cook M, Soloveichik D, Winfree E, Bruck J (2009) Programmability of chemical reaction networks. In: Condon A, Harel D, Kok JN, Salomaa A, Winfree E (eds) Algorithmic bioprocesses. Springer, Berlin, pp 543–584

    Chapter  Google Scholar 

  • Cruise J, Ganesh A (2014) Probabilistic consensus via polling and majority rules. Queueing Syst 78(2):99–120

    Article  MathSciNet  Google Scholar 

  • Doerr B, Goldberg LA, Minder L, Sauerwald T, Scheideler C (2011) Stabilizing consensus with the power of two choices. In: Proceedings of the twenty-third annual ACM symposium on parallelism in algorithms and architectures, SPAA ’11. New York, NY, USA, ACM, pp 149–158

  • Draief M, Vojnovic M (2012) Convergence speed of binary interval consensus. SIAM J Control Optim 50(3):1087–1109

    Article  MathSciNet  Google Scholar 

  • Feller W (1968) An introduction to probability theory and its applications, vol 1, 3rd edn. Wiley, New York

    MATH  Google Scholar 

  • Gillespie DT (1977) Exact stochastic simulation of coupled chemical reactions. J Phys Chem 81:2340–2361

    Article  Google Scholar 

  • Hoeffding W (1963) Probability inequalities for sums of bounded random variables. J Am Stat Assoc 58:13–30

    Article  MathSciNet  Google Scholar 

  • Mertzios GB, Nikoletseas SE, Raptopoulos CL, Spirakis PG (2017) Determining majority in networks with local interactions and very small local memory. Distrib Comput 30(1):1–16

    Article  MathSciNet  Google Scholar 

  • Perron E, Vasudevan D, Vojnovic M (2009) Using three states for binary consensus on complete graphs. In: Proceedings of the 28th IEEE conference on computer communications (INFOCOM). pp 2527–2535

  • Soloveichik D, Cook M, Winfree E, Bruck J (2008) Computation with finite stochastic chemical reaction networks. Nat Comput 7:615–633

    Article  MathSciNet  Google Scholar 

  • Soloveichik D, Seelig G, Winfree E (2010) DNA as a universal substrate for chemical kinetics. PNAS 107(12):5393–5398

    Article  Google Scholar 

  • van Kampen N (1997) Stochastic processes in physics and chemistry (revised edition)

Download references

Acknowledgements

We thank Frederik Mallman-Trenn for helpful discussions about the literature on synchronous models and their relationship to our asychronous models.

Funding

Funding for this work was provided by the Natural Sciences and Engineering Research Council of Canada’s Discovery Grants Program.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Anne Condon.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Condon, A., Hajiaghayi, M., Kirkpatrick, D. et al. Approximate majority analyses using tri-molecular chemical reaction networks. Nat Comput 19, 249–270 (2020). https://doi.org/10.1007/s11047-019-09756-4

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11047-019-09756-4

Keywords

Navigation