‘Go with the winners’ generators with applications to molecular modeling

  • Marcus Peinado
  • Thomas Lengauer
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1269)


A generation problem is the problem of generating an element of a (usually exponentially large) set under a given distribution. We develop a method for the design of generation algorithms which is based on the ‘go with the winners’ algorithm of Aldous and Vazirani [AV94]. We apply the scheme to two concrete problems from computational chemistry: the generation of models of amorphous solids and of certain kinds of polymers.


Markov Chain Problem Instance Target Distribution Output Distribution Amorphous Solid 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. [AV94]
    D. Aldous and U. Vazirani. “Go with the winners” algorithms. In Proceedings 35th IEEE Symposium on the Foundations of Computer Science, 1994.Google Scholar
  2. [BS85]
    A. Berretti and A. Sokal. New monte carlo method for the self avoiding walk. Journal of Statistical Physics, 40:483–531, 1985.Google Scholar
  3. [Dob68]
    P. L. Dobrushin. The description of a random field by means of conditional probabilities and conditions of its regularity. Theory of Probability and its Applications, 8(2):197–224, 1968.Google Scholar
  4. [E1190]
    S. Elliot. Physics of Amorphous Materials. Longman, 1990.Google Scholar
  5. [GJ79]
    M. R. Garey and D. S. Johnson. Computers and Intractability: a guide to the theory of NP-completeness. W. H. Freeman, 1979.Google Scholar
  6. [GJ97]
    L. Goldberg and M. Jerrum. Randomly sampling molecules. In Proc. of the 8th Annual ACM-SIAM Symposium on Discrete Algorithms, 1997.Google Scholar
  7. [Gra96]
    P. Grassberger. The pruned-enriched rosenbluth method: Simulations of theta-polymers of chain length up to 1,000,000. Manuscript, 1996.Google Scholar
  8. [Hen88]
    B. Hendrickson. Conditions for unique graph embeddings. Technical Report 88–950, Cornell University, Department of Computer Science, 1988.Google Scholar
  9. [JVV86]
    M. Jerrum, L. Valiant, and V. Vazirani. Random generation of combinatorial structures from a uniform distribution. Theoretical Computer Science, 43:169–188, 1986.Google Scholar
  10. [Lev73]
    L. A. Levin. Universal sorting problems. Problemy Peredachi Informatsii, 9(3):265–266, 1973. In Russian.Google Scholar
  11. [Lev86]
    L. A. Levin. Average case complete problems. SIAM Journal on Computing, 15:285–286, 1986.Google Scholar
  12. [RS94]
    Dana Randall and Alistair Sinclair. Testable algorithms for self-avoiding walks. In Proceedings of the Fifth Annual ACM-SIAM Symposium on Discrete Algorithms, pages 593–602, Philadelphia, Pennsylvania, 23–25 January 1994.Google Scholar
  13. [Sax79]
    J. Saxe. Two papers on graph embedding problems. Technical Report CMUCS-80-102, Carnegie-Mellon University, Department of Computer Science, 1979.Google Scholar
  14. [Shi84]
    A. N. Shiryayev. Probability. Springer-Verlag, 1984.Google Scholar
  15. [Sin93]
    A. Sinclair. Algorithms For Random Generation And Counting. Progress In Theoretical Computer Science. Birkhauser, Boston, 1993.Google Scholar
  16. [SJ89]
    A. Sinclair and M. Jerrum. Approximate counting, uniform generation and rapidly mixing Markov chains. Information and Computation, 82:93–133, 1989.Google Scholar
  17. [Wel93]
    D. J. A. Welsh. Complexity: Knots, Colourings and Knots. Cambridge University Press, 1993.Google Scholar
  18. [Yem79]
    Y. Yemini. Some theoretical aspects of position-location problems. In Proceedings 20th IEEE Symposium on the Foundations of Computer Science, 1979.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • Marcus Peinado
    • 1
  • Thomas Lengauer
    • 1
  1. 1.Institute for Algorithms and Scientific ComputingGerman National Research Center for Information Technology (GMD)Sankt AugustinGermany

Personalised recommendations