Limits...
Using targeted enrichment of nuclear genes to increase phylogenetic resolution in the neotropical rain forest genus Inga (Leguminosae: Mimosoideae).

Nicholls JA, Pennington RT, Koenen EJ, Hughes CE, Hearn J, Bunnefeld L, Dexter KG, Stone GN, Kidner CA - Front Plant Sci (2015)

Bottom Line: Bayesian phylogenies reconstructed using either all loci concatenated or a gene-tree/species-tree approach yielded highly resolved phylogenies.We used coalescent approaches to show that the same targeted enrichment data also have significant power to discriminate among alternative within-species population histories within the widespread species I. umbellifera.In either application, targeted enrichment simplifies the informatics challenge of identifying orthologous loci associated with de novo genome sequencing.

View Article: PubMed Central - PubMed

Affiliation: Ashworth Labs, Institute of Evolutionary Biology, School of Biological Sciences, University of Edinburgh Edinburgh, UK ; Royal Botanic Garden Edinburgh Edinburgh, UK.

ABSTRACT
Evolutionary radiations are prominent and pervasive across many plant lineages in diverse geographical and ecological settings; in neotropical rainforests there is growing evidence suggesting that a significant fraction of species richness is the result of recent radiations. Understanding the evolutionary trajectories and mechanisms underlying these radiations demands much greater phylogenetic resolution than is currently available for these groups. The neotropical tree genus Inga (Leguminosae) is a good example, with ~300 extant species and a crown age of 2-10 MY, yet over 6 kb of plastid and nuclear DNA sequence data gives only poor phylogenetic resolution among species. Here we explore the use of larger-scale nuclear gene data obtained though targeted enrichment to increase phylogenetic resolution within Inga. Transcriptome data from three Inga species were used to select 264 nuclear loci for targeted enrichment and sequencing. Following quality control to remove probable paralogs from these sequence data, the final dataset comprised 259,313 bases from 194 loci for 24 accessions representing 22 Inga species and an outgroup (Zygia). Bayesian phylogenies reconstructed using either all loci concatenated or a gene-tree/species-tree approach yielded highly resolved phylogenies. We used coalescent approaches to show that the same targeted enrichment data also have significant power to discriminate among alternative within-species population histories within the widespread species I. umbellifera. In either application, targeted enrichment simplifies the informatics challenge of identifying orthologous loci associated with de novo genome sequencing. We conclude that targeted enrichment provides the large volumes of phylogenetically-informative sequence data required to resolve relationships within recent plant species radiations, both at the species level and for within-species phylogeographic studies.

No MeSH data available.


(A) Schematic of the four different population models tested using 168 loci obtained through targeted enrichment in I. umbellifera. Codes for populations in the two different test sets are: Panama, BCI; French Guiana, FG; Ecuador, EC; and Peru, PE. Diagram modified from (Lohse et al., 2012). (B) The expected difference in support (ΔlnL) between a full (three-population) model and each nested model as a function of the number of loci for older (left panel) and more recent (right panel) scenarios. The two horizontal black lines show the ΔlnL needed to reject the simpler model at α = 0.05 when the simpler model has one fewer parameter (polytomy and two-population models, lower line) or two fewer parameters (panmixis, upper line).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4584976&req=5

Figure 5: (A) Schematic of the four different population models tested using 168 loci obtained through targeted enrichment in I. umbellifera. Codes for populations in the two different test sets are: Panama, BCI; French Guiana, FG; Ecuador, EC; and Peru, PE. Diagram modified from (Lohse et al., 2012). (B) The expected difference in support (ΔlnL) between a full (three-population) model and each nested model as a function of the number of loci for older (left panel) and more recent (right panel) scenarios. The two horizontal black lines show the ΔlnL needed to reject the simpler model at α = 0.05 when the simpler model has one fewer parameter (polytomy and two-population models, lower line) or two fewer parameters (panmixis, upper line).

Mentions: The power analyses showed fewer loci are required to resolve deeper nodes and older population histories. For the older divergence scenario (values of T1 = 0.1 and T2 = 0.4) data for fewer than 50 loci are required to resolve the older population split (T2) for either value of θ, and hence reject the two simplest unresolved population models (panmixis or a polytomy; Figure 5B, left panel). However, we would need data for almost 650/300 loci (for lower/higher θ values, respectively) to resolve the more recent population split (T1) and reject the two-population model in favor of the true three-population model. For the more recent divergence history (T1 = 0.01 and T2 = 0.1) even rejecting the simplest models in favor of a two-population model requires more loci than we currently have for I. umbellifera: we could confidently reject panmixis with 407/197 loci (lower/higher θ) or a polytomy model with 437/243 loci (lower/higher θ) (Figure 5B, right panel). We would need many thousands of loci to reject a two-population model in favor of the correct fully-resolved three-population model.


Using targeted enrichment of nuclear genes to increase phylogenetic resolution in the neotropical rain forest genus Inga (Leguminosae: Mimosoideae).

Nicholls JA, Pennington RT, Koenen EJ, Hughes CE, Hearn J, Bunnefeld L, Dexter KG, Stone GN, Kidner CA - Front Plant Sci (2015)

(A) Schematic of the four different population models tested using 168 loci obtained through targeted enrichment in I. umbellifera. Codes for populations in the two different test sets are: Panama, BCI; French Guiana, FG; Ecuador, EC; and Peru, PE. Diagram modified from (Lohse et al., 2012). (B) The expected difference in support (ΔlnL) between a full (three-population) model and each nested model as a function of the number of loci for older (left panel) and more recent (right panel) scenarios. The two horizontal black lines show the ΔlnL needed to reject the simpler model at α = 0.05 when the simpler model has one fewer parameter (polytomy and two-population models, lower line) or two fewer parameters (panmixis, upper line).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4584976&req=5

Figure 5: (A) Schematic of the four different population models tested using 168 loci obtained through targeted enrichment in I. umbellifera. Codes for populations in the two different test sets are: Panama, BCI; French Guiana, FG; Ecuador, EC; and Peru, PE. Diagram modified from (Lohse et al., 2012). (B) The expected difference in support (ΔlnL) between a full (three-population) model and each nested model as a function of the number of loci for older (left panel) and more recent (right panel) scenarios. The two horizontal black lines show the ΔlnL needed to reject the simpler model at α = 0.05 when the simpler model has one fewer parameter (polytomy and two-population models, lower line) or two fewer parameters (panmixis, upper line).
Mentions: The power analyses showed fewer loci are required to resolve deeper nodes and older population histories. For the older divergence scenario (values of T1 = 0.1 and T2 = 0.4) data for fewer than 50 loci are required to resolve the older population split (T2) for either value of θ, and hence reject the two simplest unresolved population models (panmixis or a polytomy; Figure 5B, left panel). However, we would need data for almost 650/300 loci (for lower/higher θ values, respectively) to resolve the more recent population split (T1) and reject the two-population model in favor of the true three-population model. For the more recent divergence history (T1 = 0.01 and T2 = 0.1) even rejecting the simplest models in favor of a two-population model requires more loci than we currently have for I. umbellifera: we could confidently reject panmixis with 407/197 loci (lower/higher θ) or a polytomy model with 437/243 loci (lower/higher θ) (Figure 5B, right panel). We would need many thousands of loci to reject a two-population model in favor of the correct fully-resolved three-population model.

Bottom Line: Bayesian phylogenies reconstructed using either all loci concatenated or a gene-tree/species-tree approach yielded highly resolved phylogenies.We used coalescent approaches to show that the same targeted enrichment data also have significant power to discriminate among alternative within-species population histories within the widespread species I. umbellifera.In either application, targeted enrichment simplifies the informatics challenge of identifying orthologous loci associated with de novo genome sequencing.

View Article: PubMed Central - PubMed

Affiliation: Ashworth Labs, Institute of Evolutionary Biology, School of Biological Sciences, University of Edinburgh Edinburgh, UK ; Royal Botanic Garden Edinburgh Edinburgh, UK.

ABSTRACT
Evolutionary radiations are prominent and pervasive across many plant lineages in diverse geographical and ecological settings; in neotropical rainforests there is growing evidence suggesting that a significant fraction of species richness is the result of recent radiations. Understanding the evolutionary trajectories and mechanisms underlying these radiations demands much greater phylogenetic resolution than is currently available for these groups. The neotropical tree genus Inga (Leguminosae) is a good example, with ~300 extant species and a crown age of 2-10 MY, yet over 6 kb of plastid and nuclear DNA sequence data gives only poor phylogenetic resolution among species. Here we explore the use of larger-scale nuclear gene data obtained though targeted enrichment to increase phylogenetic resolution within Inga. Transcriptome data from three Inga species were used to select 264 nuclear loci for targeted enrichment and sequencing. Following quality control to remove probable paralogs from these sequence data, the final dataset comprised 259,313 bases from 194 loci for 24 accessions representing 22 Inga species and an outgroup (Zygia). Bayesian phylogenies reconstructed using either all loci concatenated or a gene-tree/species-tree approach yielded highly resolved phylogenies. We used coalescent approaches to show that the same targeted enrichment data also have significant power to discriminate among alternative within-species population histories within the widespread species I. umbellifera. In either application, targeted enrichment simplifies the informatics challenge of identifying orthologous loci associated with de novo genome sequencing. We conclude that targeted enrichment provides the large volumes of phylogenetically-informative sequence data required to resolve relationships within recent plant species radiations, both at the species level and for within-species phylogeographic studies.

No MeSH data available.