Skip to main content

Average Fitness Differences on NK Landscapes

Theory in Biosciences volume 139, pages 1–7 (2020)Cite this article

Abstract

The average fitness difference between adjacent sites in a fitness landscape is an important descriptor that impacts in particular the dynamics of selection/mutation processes on the landscape. Of particular interest is its connection to the error threshold phenomenon. We show here that this parameter is intimately tied to the ruggedness through the landscape’s amplitude spectrum. For the NK model, a surprisingly simple analytical estimate explains simulation data with high precision.

This is a preview of subscription content, access via your institution.

Access options

Buy single article

Instant access to the full article PDF.

USD 39.95

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

Fig. 1
Fig. 2

References

  • Bonhoeffer S, Stadler PF (1993) Errorthreshold on complex fitness landscapes. J Theor Biol 164:359–372

    Article  Google Scholar 

  • Bull JJ, Ancel Meyers L, Lachmann M (2005) Quasispecies made simple. PLoS Comput Biol 1:e61

    CAS  Article  Google Scholar 

  • Buzas J, Dinitz J (2014) An analysis of NK landscapes: Interaction structure, statistical properties, and expected number of local optima. IEEE Trans Evol Comput 18:807–818

    Article  Google Scholar 

  • Campos PRA, Adami C, Wilke CO (2002) Optimal adaptive performance and delocalization in NK fitness landscapes. Physica A 304:495–506

    Article  Google Scholar 

  • de Oliveira VM, Fontanari JF, Stadler PF (1999) Metastable states in high order short-range spin glasses. J Phys A Math Gen 32:8793–8802

    Article  Google Scholar 

  • Eigen M (1971) Selforganization of matter and the evolution of biological macromolecules. Naturwissenschaften 10:465–523

    Article  Google Scholar 

  • Eigen M, McCaskill J, Schuster P (1989) The molecular quasispecies. Adv Chem Phys 75:149–263

    CAS  Google Scholar 

  • García-Pelayo R, Stadler PF (1997) Correlation length, isotropy, and meta-stable states. Physica D 107:240–254

    Article  Google Scholar 

  • Grover LK (1992) Local search and the local structure of NP-complete problems. Oper Res Lett 12:235–243

    Article  Google Scholar 

  • Hordijk W, Stadler PF (1998) Amplitude spectra of fitness landscapes. J Complex Syst 1:39–66

    Article  Google Scholar 

  • Kallel L, Naudts B, Reeves CR (2001) Properties of fitness functions and search landscapes. In: Kallel L, Naudts B, Rogers A (eds) Theoretical aspects of evolutionary computing. Springer, Heidelberg, pp 175–206

    Chapter  Google Scholar 

  • Kauffman SA (1989) Adaptation on rugged fitness landscapes. In: Stein D (ed) Lectures in the sciences of complexity. Addison-Wesley, Boston, pp 527–618

    Google Scholar 

  • Kauffman SA (1993) Origins of order: self-organization and selection in evolution. Oxford University Press, Oxford

    Google Scholar 

  • Kauffman SA, Levin S (1987) Towards a general theory of adaptive walks on rugged landscapes. J Theor Biol 128:11–45

    CAS  Article  Google Scholar 

  • Kaul H, Jacobson SH (2006) Global optima results for the kauffman NK model. Math Program 106:319–338

    Article  Google Scholar 

  • Klemm K, Stadler PF (2014) Rugged and elementary landscapes. In: Borenstein Y, Moraglio A (eds) Theory and principled methods for designing metaheustics, natural computing series. Springer, Berlin, pp 41–61

    Chapter  Google Scholar 

  • Kondrashov DA, Kondrashov FA (2015) Topological features of rugged fitness landscapes in sequence space. Trends Genet 31:24–33

    CAS  Article  Google Scholar 

  • McCaskill JS (1984) A localisation threshhold for macromolecular quasispecies from continuously distributed replication rates. J Chem Phys 80:5194–5202

    CAS  Article  Google Scholar 

  • Mohar B (1997) Some applications of Laplace eigenvalues of graphs. In: Hahn G, Sabidussi G (eds) Graph symmetry: algebraic methods and applications, vol. 497 of NATO ASI Series C. Kluwer, Dordrecht, pp 227–275

    Google Scholar 

  • Neidhart J, Szendro IG, Krug J (2013) Exact results for amplitude spectra of fitness landscapes. J Theor Biol 332:218–227

    Article  Google Scholar 

  • Nowak S, Krug J (2015) Analysis of adaptive walks on NK fitness landscapes with different interaction schemes. J Stat Mech

  • Ochoa G (2006) Error thresholds in genetic algorithms. Evol Comp 14:157–182

    Article  Google Scholar 

  • Reidys CM, Stadler PF (2001) Neutrality in fitness landscapes. Appl Math Comput 117:321–350

    Google Scholar 

  • Reidys CM, Stadler PF (2002) Combinatorial landscapes. SIAM Rev 44:3–54

    Article  Google Scholar 

  • Richter H, Engelbrecht A, editors (2014) Recent advances in the theory and application of fitness landscapes, vol. 6 of emergence, complexity, and computation. Springer, Berlin

  • Schuster P (2016) Quasispecies on fitness landscapes. In: Domingo E, Schuster P (eds) Quasispecies: from theory to experimental systems. Springer, Heidelberg, pp 61–120

    Google Scholar 

  • Semenov YS, Bratus AS, Novozhilov AS (2014) On the behavior of the leading eigenvalue of Eigen’s evolutionary matrices. Math Biosci 258:134–147

    Article  Google Scholar 

  • Stadler BMR, Stadler PF (2002) Generalized topological spaces in evolutionary theory and combinatorial chemistry. J Chem Inf Comput Sci 42:577–585

    CAS  Article  Google Scholar 

  • Stadler PF (1994) Linear operators on correlated landscapes. J Phys I France 4:681–696

    Article  Google Scholar 

  • Stadler PF (1996) Landscapes and their correlation functions. J Math Chem 20:1–45

    CAS  Article  Google Scholar 

  • Stadler PF, Happel R (1999) Random field models for fitness landscapes. J Math Biol 38:435–478

    Article  Google Scholar 

  • Walsh JL (1923) A closed set of normal orthogonal functions. Amer J Math 45:5–24

    Article  Google Scholar 

  • Weinberger E (1991) Local properties of Kauffman’s N-k model: a tunably rugged energy landscape. Phys Rev A 44:6399–6413

    CAS  Article  Google Scholar 

  • Whitley LD, Sutton AM, Howe AE (2008) Understanding elementary landscapes. In: Ryan C, Keijzer M (eds) Genetic and evolutionary computation conference, GECCO 2008, pp 585–592. ACM

  • Wiehe T (1997) Model dependency of error thresholds: the role of fitness functions and contrasts between the finite and infinite sites models. Genet Res 69:127–136

    Article  Google Scholar 

  • Wright S (1932) The roles of mutation, inbreeding, crossbreeeding and selection in evolution. In: Jones DF (ed) Proceedings of the sixth international congress on genetics, vol 1. Brooklyn Botanic Gardens, New York, pp 356–366

  • Wright S (1967) “Surfaces” of selective value. Proc Nat Acad Sci USA 58:165–172

    CAS  Article  Google Scholar 

Download references

Acknowledgements

WH thanks the Institute for Advanced Study, Amsterdam, for financial support through a fellowship. Discussions with Bärbel M. R. Stadler are gratefully acknowledged.

Author information

Authors and Affiliations

  1. SmartAnalytiX.com, Lausanne, Switzerland

    Wim Hordijk

  2. Institute for Systems Biology, Seattle, WA, USA

    Stuart A. Kauffman

  3. Bioinformatics Group, Department of Computer Science, Leipzig University, Härtelstraße 16-18, 04107, Leipzig, Germany

    Peter F. Stadler

  4. Interdisciplinary Center of Bioinformatics, Leipzig University, Härtelstraße 16-18, 04107, Leipzig, Germany

    Peter F. Stadler

  5. German Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig, Leipzig University, Härtelstraße 16-18, 04107, Leipzig, Germany

    Peter F. Stadler

  6. Competence Center for Scalable Data Services and Solutions, Leipzig University, Härtelstraße 16-18, 04107, Leipzig, Germany

    Peter F. Stadler

  7. Leipzig Research Center for Civilization Diseases, Leipzig University, Härtelstraße 16-18, 04107, Leipzig, Germany

    Peter F. Stadler

  8. Max Planck Institute for Mathematics in the Science, Inselstraße 22, 04103, Leipzig, Germany

    Peter F. Stadler

  9. University of Vienna, Institute for Theoretical Chemistry, Währingerstraße 17, 1090, Vienna, Austria

    Peter F. Stadler

  10. Facultad de Ciencias, Universidad Nacional de Colombia, Bogotá, DC, Colombia

    Peter F. Stadler

Authors
  1. Wim Hordijk
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Stuart A. Kauffman
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Peter F. Stadler
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Wim Hordijk.

Additional information

In memoriam Manfred Eigen $$^*$$∗9 May 1927–$$^\dagger $$† 6 Feb 2019.

Publisher's Note

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

Appendices

An example of an NK landscape

We include here a brief example of an NK landscape, taken from (Kauffman 1993). The fitness contributions \(f_i(b)\) for the individual positions are tabulated for each value of the bit \(b_i\) itself and \(K=2\) additional relevant bits. The final fitness value for each bit string is the average of the position-wise contributions. Together with the adjacency relation of the Boolean hypercube, this defines the landscape (Fig. 3).

Fig. 3
figure 3

A simple example of an instance of the NK model for \(N=3\) and \(K=2\). Top: The fitness contributions for the three bits for each of the \(2^{K+1}=8\) possible neighborhood configurations are assigned at random. The fitness of the entire string is the average of the individual fitness contributions. Bottom: The boolean hypercube representing the fitness landscape defined in the table above

Full size image

Independence of Parity and Neighborhood Structure

Figure 4 shows a scatter plot of \((\varDelta f)(x)/N\) against f for landscapes with different parity and different neighborhood structures (random and adjacent). We have chosen the different values of K such that the data sets are distinguishable. The predicted slopes for \(N=50\) and \(K=5\), 10, 15, and 20 are \(s=0.12\), 0.22, 0.32, and 0.42, while the empirical values from the data displayed here are \(\hat{s}=0.127\), 0.230, 0.328, and 0.412, respectively. The empirical and theoretical values are in excellent agreement.

Fig. 4
figure 4

Scatter plot of \((\varDelta f)(x)/N\) against f, illustrating that the slopes do not depend on parity, i.e., even vs. odd values of K, or the choice of adjacency, i.e., random (’rnd’) vs. adjacent (’adj’)

Full size image

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Hordijk, W., Kauffman, S.A. & Stadler, P.F. Average Fitness Differences on NK Landscapes. Theory Biosci. 139, 1–7 (2020). https://doi.org/10.1007/s12064-019-00296-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s12064-019-00296-0

Keywords

  • NK model
  • Fitness landscapes
  • Elementary landscape
  • Graph Laplacian
  • Error threshold