Defining Coalescent Genes: Theory Meets Practice in Organelle Phylogenomics. (30th June 2021)

• Record Type: Journal Article
• Authors: Doyle, Jeff J
• Editors: Carstens, Bryan
• Is Part Of: Systematic biology. Volume 71:Number 2(2022)
• Page Start: 476
• Page End: 489
• Publication Date: 2021-06-30
• Subjects: Biology -- Classification -- Periodicals; Biology -- Periodicals; Biologie -- Classification -- Périodiques; Biologie -- Périodiques; 578.012
• DOI: 10.1093/sysbio/syab053
• Languages: English
• ISSNs: 1063-5157

Abstract

Abstract: The species tree paradigm that dominates current molecular systematic practice infers species trees from collections of sequences under assumptions of the multispecies coalescent (MSC), that is, that there is free recombination between the sequences and no (or very low) recombination within them. These coalescent genes (c-genes) are thus defined in an historical rather than molecular sense and can in theory be as large as an entire genome or as small as a single nucleotide. A debate about how to define c-genes centers on the contention that nuclear gene sequences used in many coalescent analyses undergo too much recombination, such that their introns comprise multiple c-genes, violating a key assumption of the MSC. Recently a similar argument has been made for the genes of plastid (e.g., chloroplast) and mitochondrial genomes, which for the last 30 or more years have been considered to represent a single c-gene for the purposes of phylogeny reconstruction because they are nonrecombining in an historical sense. Consequently, it has been suggested that these genomes should be analyzed using coalescent methods that treat their genes—over 70 protein-coding genes in the case of most plastid genomes (plastomes)—as independent estimates of species phylogeny, in contrast to the usual practice of concatenation, which is appropriate for generating gene trees. However, although recombination certainly occurs in the plastome, as has been recognized since the 1970's, it is unlikely to be phylogenetically relevant. This is because such historically effective recombination can only occur when plastomes with incongruent histories are brought together in the same plastid. However, plastids sort rapidly into different cell lineages and rarely fuse. Thus, because of plastid biology, the plastome is a more canonical c-gene than is the average multi-intron mammalian nuclear gene. The plastome should thus continue to be treated as a single estimate of the underlying species phylogeny, as should the mitochondrial genome. The implications of this long-held insight of molecular systematics for studies in the phylogenomic era are explored. [c-gene; coalescent gene; concatalescence; organelle genome; plastome; recombination; species tree.]