PAReTT: A Python Package for the Automated Retrieval and Management of Divergence Time Data from the TimeTree Resource for Downstream Analyses

Abstract

Evolutionary processes happen gradually over time and are, thus, considered time dependent. In addition, several evolutionary processes are either adaptations to local habitats or changing habitats, otherwise restricted thereby. Since evolutionary processes driving speciation take place within the landscape of environmental and temporal bounds, several published studies have aimed at providing accurate, fossil-calibrated, estimates of the divergence times of both extant and extinct species. Correct calibration is critical towards attributing evolutionary adaptations and speciation both to the time and paleogeography that contributed to it. Data from more than 4000 studies and nearly 1,50,000 species are available from a central TimeTree resource and provide opportunities of retrieving divergence times, evolutionary timelines, and time trees in various formats for most vertebrates. These data greatly enhance the ability of researchers to investigate evolution. However, there is limited functionality when studying lists of species that require batch retrieval. To overcome this, a PYTHON package termed Python-Automated Retrieval of TimeTree data (PAReTT) was created to facilitate a biologist-friendly interaction with the TimeTree resource. Here, we illustrate the use of the package through three examples that includes the use of timeline data, time tree data, and divergence time data. Furthermore, PAReTT was previously used in a meta-analysis of candidate genes to illustrate the relationship between divergence times and candidate genes of migration. The PAReTT package is available for download from GitHub or as a pre-compiled Windows executable, with extensive documentation on the package available on GitHub wiki pages regarding dependencies, installation, and implementation of the various functions.

Keywords: Divergence time; Diversification rate; PAReTT; PYTHON; Time trees; Timelines.

Fig. 1

Paleogeographic reconstructions of Earth for the past 70 million years to illustrate environmental differences during the period of divergence using PALEOMAP. a Positions of the continents approximately 70 million years ago during the Upper Cretaceous before West Africa had merged with the main continent and when India was still an island. b Continents approximately 50 million years ago during the Palaeocene after the African continent formed but before the Americas were connected and shortly before India merged with Asia or the polar caps formed. c Geography of Earth during the Eocene, 30 million years ago, by which time most continents had formed but central America did not connect the North and South yet and much of Europe was still under water. d Modern day geography of Earth in the current Holocene. (Image created in BioRender.com)

Fig. 2

Graphical summary of menu options and submenus for PAReTT. The main menu options are: *, to verify the availability of data; a or b to resolve divergence times (pair or batch); c retrieve a timeline; d retrieve a time tree; e, print the TimeTree citation; f validate data; or q to exit. The data availability, timeline, and time tree options bring up a submenu, indicated by yellow arrows, to retrieve information for an individual species/taxon or a list. Output generated for lists are exported as a table (CSV), images (JPEG), or trees (Newick). The validation option brings up the choice of finding or replacing missing values for divergence times data (options a or b) as well as to view the tree topology (option c), for output files. (Image created in BioRender.com)

Fig. 3

Example of an evolutionary timeline (left) retrieved using PAReTT for the Lazuli bunting. The left panel of the timeline indicates the major geologic timescale of the past 2000 MYA including the prevailing Phanerozoic Eon as well as the subdivisions by era Paleozoic which lasted until approximately 250 MYA when the Mesozoic started which lasted until 65 MYA. As is illustrated, this period was marked by the emergence of the first birds in the class Aves approximately 110 MYA. The species currently recognized in the family Cardinalidae emerged in the Cenozoic era approximately 11 MYA, with most species of the genus Passerina first emerging 4 MYA. On the right and in the middle, the paleogeography of the North American continent during this period is illustrated, clearly showing the separation of the Western and Eastern parts by the Western interior seaway (Sampson et al. 2010). Next to this is the contemporary range map for the species showing their breeding and wintering grounds that still encompass most of the Western half of the continent (Greene et al. 2020). (Image created in BioRender.com)

Fig. 4

Comparison of time trees retrieved for two genera using PAReTT, one for several species in the Catharus genus of Neotropical thrushes and another of species in the Malurus genus of Australasian fairywrens. The Catharus genus has 12 species, including Swainson’s thrush, which diverged over a period of 4.74 MYA with a diversification rate of 0.38. The Malurus genus has 11 species, including the Superb fairywren, which diverged over a period of 9 MYA with a diversification rate of 0.19. This illustrates how two lineages may have similar evolutionary processes happening but have different diversification rates due to the difference in time scales. (Image created in BioRender.com)

Fig. 5

Plots from the PyRate analysis indicating the A speciation, B extinction, and C diversification rates over time for two genera of birds based on node age data from TimeTree. The left panel is the results for the Catharus thrush species complex while the right indicated the results for the Malurus fairywren species complex. Both genera followed similar patterns of speciation and diversification; however, the timescales over which these processes occurred are different and the Catharus species has slightly higher rates. As no extinct species could be included, the extinction rates (μ) are zero, resulting in highly similar speciation (λ) and diversification (r) plots. (Image created in BioRender.com)

Fig. 6

Mantel correlograms from the comparison of distance as measured by divergence times to genetic distance for gene 1 (A) and gene 2 (B). Divergence times were retrieved for forty bird species using PAReTT while genetic distance was measured by computing the fixation index (F_ST) for two individual genes. For both genes, a significant correlation was found between the genetic distance and divergence times where increases in genetic distance corresponded to older divergence times. (*p-value < 0.1, **p-value < 0.05)