Getting insight into the pan-genome structure with PangTree

Abstract

Background: The term pan-genome was proposed to denominate collections of genomic sequences jointly analyzed or used as a reference. The constant growth of genomic data intensifies development of data structures and algorithms to investigate pan-genomes efficiently. Results: This work focuses on providing a tool for discovering and visualizing the relationships between the sequences constituting a pan-genome. A new structure to represent such relationships - called affinity tree - is proposed. Each node of this tree has assigned a subset of genomes, as well as their homogeneity level and averaged consensus sequence. Moreover, subsets assigned to sibling nodes form a partition of the genomes assigned to their parent. Conclusions: Functionality of affinity tree is demonstrated on simulated data and on the Ebola virus pan-genome. Furthermore, two software packages are provided: PangTreeBuild constructs affinity tree, while PangTreeVis presents its result.