Computational complexity of inferring phylogenies from dissimilarity matrices (Q1091978)

Molecular biologists strive to infer evolutionary relationships from quantitative macromolecular comparisons obtained by immunological, DNA hybridization, electrophoretic or amino acid sequencing techniques. The problem is to find unrooted phylogenies that best approximate a given dissimilarity matrix according to a goodness-of-fit measure, for example the least-squares-fit criterion or Farris's f statistic [Am. Nat. 106, 645-668 (1972)]. Computational costs of known algorithms guaranteeing optimal solutions to these problems increase exponentially with problem size; practical computational considerations limit the algorithms to analyzing small problems. It is established here that problems of phylogenetic inference based on the least-squares-fit criterion and the f statistic are NP-complete and thus are so difficult computationally that efficient optimal algorithms are unlikely to exist for them.