Bacterial phylogenetic reconstruction from whole genomes is robust to recombination but demographic inference is not.
Phylogenetic inference in bacterial genomics is fundamental to understanding problems such as population history, antimicrobial resistance, and transmission dynamics. The field has been plagued by an apparent state of contradiction since the distorting effects of recombination on phylogeny were disc...
| Main Authors: | , |
|---|---|
| Format: | Journal article |
| Language: | English |
| Published: |
American Society for Microbiology
2014
|
| _version_ | 1797057982315364352 |
|---|---|
| author | Hedge, J Wilson, D |
| author_facet | Hedge, J Wilson, D |
| author_sort | Hedge, J |
| collection | OXFORD |
| description | Phylogenetic inference in bacterial genomics is fundamental to understanding problems such as population history, antimicrobial resistance, and transmission dynamics. The field has been plagued by an apparent state of contradiction since the distorting effects of recombination on phylogeny were discovered more than a decade ago. Researchers persist with detailed phylogenetic analyses while simultaneously acknowledging that recombination seriously misleads inference of population dynamics and selection. Here we resolve this paradox by showing that phylogenetic tree topologies based on whole genomes robustly reconstruct the clonal frame topology but that branch lengths are badly skewed. Surprisingly, removing recombining sites can exacerbate branch length distortion caused by recombination.Phylogenetic tree reconstruction is a popular approach for understanding the relatedness of bacteria in a population from differences in their genome sequences. However, bacteria frequently exchange regions of their genomes by a process called homologous recombination, which violates a fundamental assumption of phylogenetic methods. Since many researchers continue to use phylogenetics for recombining bacteria, it is important to understand how recombination affects the conclusions drawn from these analyses. We find that whole-genome sequences afford great accuracy in reconstructing evolutionary relationships despite concerns surrounding the presence of recombination, but the branch lengths of the phylogenetic tree are indeed badly distorted. Surprisingly, methods to reduce the impact of recombination on branch lengths can exacerbate the problem. |
| first_indexed | 2024-03-06T19:44:07Z |
| format | Journal article |
| id | oxford-uuid:21aebd8c-f816-4d34-a5d5-ab013861cef0 |
| institution | University of Oxford |
| language | English |
| last_indexed | 2024-03-06T19:44:07Z |
| publishDate | 2014 |
| publisher | American Society for Microbiology |
| record_format | dspace |
| spelling | oxford-uuid:21aebd8c-f816-4d34-a5d5-ab013861cef02022-03-26T11:34:43ZBacterial phylogenetic reconstruction from whole genomes is robust to recombination but demographic inference is not.Journal articlehttp://purl.org/coar/resource_type/c_dcae04bcuuid:21aebd8c-f816-4d34-a5d5-ab013861cef0EnglishSymplectic Elements at OxfordAmerican Society for Microbiology2014Hedge, JWilson, DPhylogenetic inference in bacterial genomics is fundamental to understanding problems such as population history, antimicrobial resistance, and transmission dynamics. The field has been plagued by an apparent state of contradiction since the distorting effects of recombination on phylogeny were discovered more than a decade ago. Researchers persist with detailed phylogenetic analyses while simultaneously acknowledging that recombination seriously misleads inference of population dynamics and selection. Here we resolve this paradox by showing that phylogenetic tree topologies based on whole genomes robustly reconstruct the clonal frame topology but that branch lengths are badly skewed. Surprisingly, removing recombining sites can exacerbate branch length distortion caused by recombination.Phylogenetic tree reconstruction is a popular approach for understanding the relatedness of bacteria in a population from differences in their genome sequences. However, bacteria frequently exchange regions of their genomes by a process called homologous recombination, which violates a fundamental assumption of phylogenetic methods. Since many researchers continue to use phylogenetics for recombining bacteria, it is important to understand how recombination affects the conclusions drawn from these analyses. We find that whole-genome sequences afford great accuracy in reconstructing evolutionary relationships despite concerns surrounding the presence of recombination, but the branch lengths of the phylogenetic tree are indeed badly distorted. Surprisingly, methods to reduce the impact of recombination on branch lengths can exacerbate the problem. |
| spellingShingle | Hedge, J Wilson, D Bacterial phylogenetic reconstruction from whole genomes is robust to recombination but demographic inference is not. |
| title | Bacterial phylogenetic reconstruction from whole genomes is robust to recombination but demographic inference is not. |
| title_full | Bacterial phylogenetic reconstruction from whole genomes is robust to recombination but demographic inference is not. |
| title_fullStr | Bacterial phylogenetic reconstruction from whole genomes is robust to recombination but demographic inference is not. |
| title_full_unstemmed | Bacterial phylogenetic reconstruction from whole genomes is robust to recombination but demographic inference is not. |
| title_short | Bacterial phylogenetic reconstruction from whole genomes is robust to recombination but demographic inference is not. |
| title_sort | bacterial phylogenetic reconstruction from whole genomes is robust to recombination but demographic inference is not |
| work_keys_str_mv | AT hedgej bacterialphylogeneticreconstructionfromwholegenomesisrobusttorecombinationbutdemographicinferenceisnot AT wilsond bacterialphylogeneticreconstructionfromwholegenomesisrobusttorecombinationbutdemographicinferenceisnot |