• DocumentCode
    3204853
  • Title

    Bounds for parametric sequence comparison

  • Author

    Fernandez-Baca, David ; Seppalainen, Timo ; Slutzki, Giora

  • Author_Institution
    Dept. of Comput. Sci., Iowa State Univ., Ames, IA, USA
  • fYear
    1999
  • fDate
    1999
  • Firstpage
    55
  • Lastpage
    62
  • Abstract
    We consider the problem of computing a global alignment between two or more sequences subject to varying mismatch and indel penalties. We prove a tight 3(n/2π)2/3+O(n1/3logn) bound on the worst-case number of distinct optimum alignments for two sequences of length n as the parameters are varied. This refines a O(n 2/3) upper bound by D. Gusfield et al. (1994). Our lower bound requires an unbounded alphabet. For strings over a binary alphabet, we prove a Ω(n1/2) lower bound. For the parametric global alignment of k⩾2 sequences under sum-of-pairs scoring, we prove a 3((k/2)n/2π)2/3+O(k2/3n1/3logn) upper bound on the number of distinct optimality regions and a Ω(n 2/3) lower bound. Based on experimental evidence, we conjecture that for two random sequences, the number of optimality regions is approximately √n with high probability
  • Keywords
    computational complexity; sequences; set theory; string matching; theorem proving; binary alphabet; distinct optimality regions; distinct optimum alignments; global alignment; indel penalties; lower bound; optimality regions; parametric global alignment; parametric sequence comparison; probability; random sequences; sum-of-pairs scoring; unbounded alphabet; upper bound; varying mismatch; worst-case number; Books; Computer science; Mathematics; Random sequences; Upper bound;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    String Processing and Information Retrieval Symposium, 1999 and International Workshop on Groupware
  • Conference_Location
    Cancun
  • Print_ISBN
    0-7695-0268-7
  • Type

    conf

  • DOI
    10.1109/SPIRE.1999.796578
  • Filename
    796578