Abstract Delimiting species boundaries in complex speciation scenarios, including those involving gene flow and demographic fluctuations, remains a challenge, particularly given the diversity of existing species concepts. Promising recent approaches adopt an integrative taxonomy that combines multiple sources of evidence (e.g., genetic, morphology, geographic distributions), reflecting different properties associated with the dynamics of the speciation continuum. The use of statistical inference methods for model comparison, such as approximate Bayesian computation, approximate likelihood approaches, and machine learning, has improved the assessment of species boundaries in such contexts. However, most existing approaches involve analyzing genetic and phenotypic/geographical information separately, followed by visual/qualitative comparison. Methods that integrate genetic information with other sources of evidence remain limited to simple evolutionary models and are typically unable to analyze more than a few hundred loci across a maximum of a few tens of samples. Here, we present a deep learning approach (DeepID) that combines two convolutional neural networks to integrate genomic data (thousands of loci or single nucleotide polymorphisms, SNPs) and trait information into a unified framework. Using both simulated and empirical data sets, we evaluate the power and accuracy of this approach for discriminating among competing divergence speciation scenarios (with minimal ongoing gene flow) across a varying number of SNPs and traits, as well as different levels of missing data. Analyses based on genomic or trait data alone yielded slightly lower accuracy, whereas integrating genomic and trait data resulted in improved performance. When we violated the speciation model by including extensive migration, approaches incorporating trait data were less affected than those relying solely on genomic information. Moreover, our approach successfully recovered the expected delimitation scenarios in empirical data sets from a plant (Euphorbia balsamifera) and a fish (Lepomis megalotis) species complex. We argue that our method is a flexible and promising approach, allowing for the comparison of complex scenarios and the integration of multiple data types. Combining genomic and trait data likely captures complementary signals associated with different stages of the speciation process, reflecting the fact that speciation is a continuum in which genetic and phenotypic divergence may proceed at different rates.
Abstract The repeated and independent evolution of convergent forms is one of the most remarkable patterns across the Tree of Life. Convergent morphologies, commonly associated with adaptive evolution, can evolve through multiple evolutionary pathways, whose full breadth is not yet well-understood. Moreover, most evidence for convergent evolution in mammals is qualitative and/or based on small taxonomic scales. Hystricomorpha is one of the major rodent radiations and represents an outstanding opportunity to investigate the extent of convergent evolution in shaping similar morphologies at a global scale. Despite its impressive ecomorphological disparity, the contribution of convergent evolution in shaping hystricomorph morphologies has never been assessed. Here, we quantify the degree of morphofunctional convergence of the skull of Hystricomorpha rodents using a combination of geometric morphometrics and comparative phylogenetic methods, discovering multiple instances of convergence in their craniomandibular complex morphology. Moreover, we show that convergent morphologies are associated with adaptations to specific ecological habits. For this, we gathered the largest morphological dataset for this clade, comprising 3,102 individuals from 189 species, covering the major diversity of the group. The Hystricomorpha skull convergently evolved primarily in subterranean and semiaquatic species, underscoring how unique evolutionary adaptations to those extreme conditions (relative to the ancestral habit) led to convergent morphologies. More broadly, our findings provide quantitative evidence for the prevalence of convergence across large taxonomic and geographic scales in the evolutionary history of mammals and highlight the importance of examining form-function relationships at a macroevolutionary scale.
Abstract Amino-acid substitution models are central to phylogenomic inference, yet their adequacy (i.e. whether they realistically describe features of real datasets) remains insufficiently tested. Here, we evaluated the adequacy of seven widely used models of amino acid substitution across empirical datasets spanning diverse evolutionary timescales. We focused specifically on the model's ability to capture across-site compositional heterogeneity in empirical data, by evaluating complementary metrics that quantify both site-level diversity and amino-acid frequency spectrum. Of the models we tested, LG+G, representing the site-homogeneous models, always fails to adequately describe the real data. Among site-heterogeneous mixture models, Poisson+CAT+G consistently achieves a close match to empirical data, while other finite-mixture models show intermediate or mixed behavior, with performance varying by datasets and by metric. Taken together, our results highlight that adequacy tests are important for identifying model strengths and limitations in accounting for across-site compositional heterogeneity.
Probabilistic modeling frameworks are powerful tools for statistical modeling and inference. They are not immediately generalizable to phylogenetic problems due to the particular computational properties of the phylogenetic tree object. TreeFlow is a software library for probabilistic modeling and automatic differentiation with phylogenetic trees. It embeds phylogenetic trees in the TensorFlow Probability framework, and implements inference algorithms for phylogenetic models given a fixed tree topology. We demonstrate how TreeFlow can be used to quickly implement and assess new models. We also show that it provides reasonable performance for gradient-based inference algorithms compared with specialized computational libraries for phylogenetics.
Abstract Phylogenetic methods typically generate bifurcating trees reflecting monophyletic groups formed by populations. However, unexpected topologies sometimes emerge, raising questions about whether they stem from methodological artifacts or biological phenomena. In a recent transcriptome-based study of Dugesia species, asexual populations displayed an unusual phylogenetic pattern: instead of forming monophyletic groups, each individual emerged as a long-branched independent lineage, resulting in a distinctive "ladder-pattern." While such a pattern could result from methodological issues, its exclusive presence in asexual populations suggests a biological explanation—long-term asexuality. Traditionally considered evolutionary dead ends due to reduced genetic diversity and weakened purifying selection, extant long-term asexual lineages challenge these assumptions, being popularly known as "asexual scandals". Here, we test both hypotheses: first, assessing whether the ladder-pattern arises from methodological artifacts, and second, examining expected genomic signatures of prolonged asexuality or hybridization. Using transcriptomic and nuclear marker datasets, we employ state-of-the-art phylogenetic methodologies and intraindividual genetic diversity analyses. Our results confirm that the ladder-pattern is not an artifact. Comparative analyses with other fissiparous planarian lineages support the deep evolutionary history of these asexual populations. Furthermore, we demonstrate that retained ancestral polymorphisms obscure phylogenetic signals. Our findings highlight the limitations of traditional phylogenetic methods in the context of asexuality and underscore the need to renew the models traditionally used in evolutionary studies.
