Chapter

Logic for Programming, Artificial Intelligence, and Reasoning

Volume 4790 of the series Lecture Notes in Computer Science pp 423-437

Faster Phylogenetic Inference with MXG

  • David G. MitchellAffiliated withComputational Logic Laboratory, Simon Fraser University, Burnaby BC
  • , Faraz HachAffiliated withComputational Logic Laboratory, Simon Fraser University, Burnaby BC
  • , Raheleh MohebaliAffiliated withComputational Logic Laboratory, Simon Fraser University, Burnaby BC

* Final gross prices may vary according to local VAT.

Get Access

Abstract

We apply the logic-based declarative programming approach of Model Expansion (MX) to a phylogenetic inference task. We axiomatize the task in multi-sorted first-order logic with cardinality constraints. Using the model expansion solver MXG and SAT+cardinality solver MXC, we compare the performance of several MX axiomatizations on a challenging set of test instances. Our methods perform orders of magnitude faster than previously reported declarative solutions. Our best solution involves polynomial-time pre-processing, redundant axioms, and symmetry-breaking axioms. We also discuss our method of test instance generation, and the role of pre-processing in declarative programming.

Keywords

Phylogeny Declarative Programming Model Expansion