Objective functions

Haluk Doǧan, Hasan H. Otu

Research output: Chapter in Book/Report/Conference proceedingChapter

8 Scopus citations

Abstract

Multiple sequence alignment involves alignment of more than two sequences and is an NP-complete problem. Therefore, heuristic algorithms that use different criteria to find an approximation to the optimal solution are employed. At the heart of these approaches lie the scoring and objective functions that a given algorithm uses to compare competing solutions in constructing a multiple sequence alignment. These objective functions are often motivated by the biological paradigms that govern functional similarities and evolutionary relations. Most existing approaches utilize a progressive process where the final alignment is constructed sequentially by adding new sequences into an existing multiple sequence alignment matrix, which is dynamically updated. In doing this, the core scoring function to assess accuracies of pairwise alignments generally remains the same, while the objective functions used in intermediary steps differ. Nevertheless, the overall assessment of the final multiple sequence alignment is generally calculated by an extension of pairwise scorings. In this chapter, we explore different scoring and objective functions used in calculating the accuracy and optimization of a multiple sequence alignment and provide utilization of these criteria in popularly used multiple sequence alignment algorithms.

Original languageEnglish (US)
Title of host publicationMultiple Sequence Alignment Methods
PublisherHumana Press Inc.
Pages45-58
Number of pages14
ISBN (Print)9781627036450
DOIs
StatePublished - 2014

Publication series

NameMethods in Molecular Biology
Volume1079
ISSN (Print)1064-3745

Keywords

  • Consistency-based-scoring
  • Progressive-alignment
  • Scoring-functions
  • Scoring-matrices
  • Sum-of-pairs

ASJC Scopus subject areas

  • Molecular Biology
  • Genetics

Fingerprint Dive into the research topics of 'Objective functions'. Together they form a unique fingerprint.

Cite this