The genetics and evolution of eye color in domestic pigeons (Columba livia).

The eye color of birds, generally referring to the color of the iris, results from both pigmentation and structural coloration. Avian iris colors exhibit striking interspecific and intraspecific variations that correspond to unique evolutionary and ecological histories. Here, we identified the genetic basis of pearl (white) iris color in domestic pigeons (Columba livia) to explore the largely unknown genetic mechanism underlying the evolution of avian iris coloration. Using a genome-wide association study (GWAS) approach in 92 pigeons, we mapped the pearl iris trait to a 9 kb region containing the facilitative glucose transporter gene SLC2A11B. A nonsense mutation (W49X) leading to a premature stop codon in SLC2A11B was identified as the causal variant. Transcriptome analysis suggested that SLC2A11B loss of function may downregulate the xanthophore-differentiation gene CSF1R and the key pteridine biosynthesis gene GCH1, thus resulting in the pearl iris phenotype. Coalescence and phylogenetic analyses indicated that the mutation originated approximately 5,400 years ago, coinciding with the onset of pigeon domestication, while positive selection was likely associated with artificial breeding. Within Aves, potentially impaired SLC2A11B was found in six species from six distinct lineages, four of which associated with their signature brown or blue eyes and lack of pteridine. Analysis of vertebrate SLC2A11B orthologs revealed relaxed selection in the avian clade, consistent with the scenario that during and after avian divergence from the reptilian ancestor, the SLC2A11B-involved development of dermal chromatophores likely degenerated in the presence of feather coverage. Our findings provide new insight into the mechanism of avian iris color variations and the evolution of pigmentation in vertebrates.

Are Nonsynonymous Transversions Generally More Deleterious than Nonsynonymous Transitions?

It has been suggested that, due to the structure of the genetic code, nonsynonymous transitions are less likely than transversions to cause radical changes in amino acid physicochemical properties so are on average less deleterious. This view was supported by some but not all mutagenesis experiments. Because laboratory measures of fitness effects have limited sensitivities and relative frequencies of different mutations in mutagenesis studies may not match those in nature, we here revisit this issue using comparative genomics. We extend the standard codon model of sequence evolution by adding the parameter η that quantifies the ratio of the fixation probability of transitional nonsynonymous mutations to that of transversional nonsynonymous mutations. We then estimate η from the concatenated alignment of all protein-coding DNA sequences of two closely related genomes. Surprisingly, η ranges from 0.13 to 2.0 across 90 species pairs sampled from the tree of life, with 51 incidences of η < 1 and 30 incidences of η >1 that are statistically significant. Hence, whether nonsynonymous transversions are overall more deleterious than nonsynonymous transitions is species-dependent. Because the corresponding groups of amino acid replacements differ between nonsynonymous transitions and transversions, η is influenced by the relative exchangeabilities of amino acid pairs. Indeed, an extensive search reveals that the large variation in η is primarily explainable by the recently reported among-species disparity in amino acid exchangeabilities. These findings demonstrate that genome-wide nucleotide substitution patterns in coding sequences have species-specific features and are more variable among evolutionary lineages than are currently thought.

Deep Residual Neural Networks Resolve Quartet Molecular Phylogenies.

Phylogenetic inference is of fundamental importance to evolutionary as well as other fields of biology, and molecular sequences have emerged as the primary data for this task. Although many phylogenetic methods have been developed to explicitly take into account substitution models of sequence evolution, such methods could fail due to model misspecification or insufficiency, especially in the face of heterogeneities in substitution processes across sites and among lineages. In this study, we propose to infer topologies of four-taxon trees using deep residual neural networks, a machine learning approach needing no explicit modeling of the subject system and having a record of success in solving complex nonlinear inference problems. We train residual networks on simulated protein sequence data with extensive amino acid substitution heterogeneities. We show that the well-trained residual network predictors can outperform existing state-of-the-art inference methods such as the maximum likelihood method on diverse simulated test data, especially under extensive substitution heterogeneities. Reassuringly, residual network predictors generally agree with existing methods in the trees inferred from real phylogenetic data with known or widely believed topologies. Furthermore, when combined with the quartet puzzling algorithm, residual network predictors can be used to reconstruct trees with more than four taxa. We conclude that deep learning represents a powerful new approach to phylogenetic reconstruction, especially when sequences evolve via heterogeneous substitution processes. We present our best trained predictor in a freely available program named Phylogenetics by Deep Learning (PhyDL, https://gitlab.com/ztzou/phydl; last accessed January 3, 2020).

Sorting Out the Genetic Background of the Last Surviving South China Tigers.

The South China tiger (Panthera tigris amoyensis) is endemic to China and also the most critically endangered subspecies of living tigers. It is considered extinct in the wild and only about 150 individuals survive in captivity to date, whose genetic heritage, however, is ambiguous and controversial. Here, we conducted an explicit genetic assessment of 92 studbook-registered South China tigers from 14 captive facilities using a subspecies-diagnostic system in the context of comparison with other voucher specimens to evaluate the genetic ancestry and level of distinctiveness of the last surviving P. t. amoyensis. Three mtDNA haplotypes were identified from South China tigers sampled in this study, including a unique P. t. amoyensis AMO1 haplotype not found in other subspecies, a COR1 haplotype that is widespread in Indochinese tigers (P. t. corbetti), and an ALT haplotype that is characteristic of Amur tigers (P. t. altaica). Bayesian STRUCTURE analysis and parentage verification confirmed the verified subspecies ancestry (VSA) as the South China tiger in 74 individuals. Genetic introgression from other tigers was detected in 18 tigers, and subsequent exclusion of these and their offspring from the breeding program is recommended. Both STRUCTURE clustering and microsatellite-based phylogenetic analyses demonstrated a close genetic association of the VSA South China tigers to Indochinese tigers, an issue that could only be elucidated by analysis of historical South China tiger specimens with wild origin. Our results also indicated a moderate level of genetic diversity in the captive South China tiger population, suggesting a potential for genetic restoration.

Gene Tree Discordance Does Not Explain Away the Temporal Decline of Convergence in Mammalian Protein Sequence Evolution.

Several authors reported lower frequencies of protein sequence convergence between more distantly related evolutionary lineages and attributed this trend to epistasis, which renders the acceptable amino acids at a site more different and convergence less likely in more divergent lineages. A recent primate study, however, suggested that this trend is at least partially and potentially entirely an artifact of gene tree discordance (GTD). Here, we demonstrate in a genome-wide data set from 17 mammals that the temporal trend remains (1) upon the control of the GTD level, (2) in genes whose genealogies are concordant with the species tree, and (3) for convergent changes, which are extremely unlikely to be caused by GTD. Similar results are observed in a comparable data set of 12 fruit flies in some but not all of these tests. We conclude that, at least in some cases, the temporal decline of convergence is genuine, reflecting an impact of epistasis on protein evolution.

Are Convergent and Parallel Amino Acid Substitutions in Protein Evolution More Prevalent Than Neutral Expectations?

Convergent and parallel amino acid substitutions in protein evolution, collectively referred to as molecular convergence here, have small probabilities under neutral evolution. For this reason, molecular convergence is commonly viewed as evidence for similar adaptations of different species. The surge in the number of reports of molecular convergence in the last decade raises the intriguing question of whether molecular convergence occurs substantially more frequently than expected under neutral evolution. We here address this question using all one-to-one orthologous proteins encoded by the genomes of 12 fruit fly species and those encoded by 17 mammals. We found that the expected amount of molecular convergence varies greatly depending on the specific neutral substitution model assumed at each amino acid site and that the observed amount of molecular convergence is explainable by neutral models incorporating site-specific information of acceptable amino acids. Interestingly, the total number of convergent and parallel substitutions between two lineages, relative to the neutral expectation, decreases with the genetic distance between the two lineages, regardless of the model used in computing the neutral expectation. We hypothesize that this trend results from differences in the amino acids acceptable at a given site among different clades of a phylogeny, due to prevalent epistasis, and provide simulation as well as empirical evidence for this hypothesis. Together, our study finds no genomic evidence for higher-than-neutral levels of molecular convergence, but suggests the presence of abundant epistasis that decreases the likelihood of molecular convergence between distantly related lineages.

No genome-wide protein sequence convergence for echolocation.

Toothed whales and two groups of bats independently acquired echolocation, the ability to locate and identify objects by reflected sound. Echolocation requires physiologically complex and coordinated vocal, auditory, and neural functions, but the molecular basis of the capacity for echolocation is not well understood. A recent study suggested that convergent amino acid substitutions widespread in the proteins of echolocators underlay the convergent origins of mammalian echolocation. Here, we show that genomic signatures of molecular convergence between echolocating lineages are generally no stronger than those between echolocating and comparable nonecholocating lineages. The same is true for the group of 29 hearing-related proteins claimed to be enriched with molecular convergence. Reexamining the previous selection test reveals several flaws and invalidates the asserted evidence for adaptive convergence. Together, these findings indicate that the reported genomic signatures of convergence largely reflect the background level of sequence convergence unrelated to the origins of echolocation.

Two chromosome-level genome assemblies of galling aphids Slavum lentiscoides and Chaetogeoica ovagalla.

Slavum lentiscoides and Chaetogeoica ovagalla are two aphid species from the subtribe Fordina of Fordini within the subfamily Eriosomatinae, and they produce galls on their primary host plants Pistacia. We assembled chromosome-level genomes of these two species using Nanopore long-read sequencing and Hi-C technology. A 332 Mb genome assembly of S. lentiscoides with a scaffold N50 of 19.77 Mb, including 11,747 genes, and a 289 Mb genome assembly of C. ovagalla with a scaffold N50 of 11.85 Mb, containing 14,492 genes, were obtained. The Benchmarking Universal Single-Copy Orthologs (BUSCO) benchmark of the two genome assemblies reached 93.7% (91.9% single-copy) and 97.0% (95.3% single-copy), respectively. The high-quality genome assemblies in our study provide valuable resources for future genomic research of galling aphids.

The evolution of conglobation in Ceratocanthinae.

Conglobation is an adaptive behaviour occurring independently in various animal groups. Here, we study the evolution of conglobation in Ceratocanthinae, a beetle group with the ability to roll three body segments into a tight ball. It is here implied that this ability evolved only once in the Mesozoic. Evidence is offered suggesting that the high defensive strength of Ceratocanthinae is due not only to the spherical body shape but also to the thickness and stronger mechanical properties of the dorsal cuticle. We further validate five adaptive characters including the allometrically thickened body wall and find that the specific adaptation of different body segments are likely separate evolutionary events. Finally, we propose an "attackers stress" hypothesis to explain the origin of conglobation behaviours. This work contributes to understanding how and why conglobation behaviour may have evolved in this group.

Idiosyncratic epistasis creates universals in mutational effects and evolutionary trajectories.

Patterns of epistasis and shapes of fitness landscapes are of wide interest because of their bearings on a number of evolutionary theories. The common phenomena of slowing fitness increases during adaptations and diminishing returns from beneficial mutations are believed to reflect a concave fitness landscape and a preponderance of negative epistasis. Paradoxically, fitness decreases tend to decelerate and harm from deleterious mutations shrinks during the accumulation of random mutations-patterns thought to indicate a convex fitness landscape and a predominance of positive epistasis. Current theories cannot resolve this apparent contradiction. Here, we show that the phenotypic effect of a mutation varies substantially depending on the specific genetic background and that this idiosyncrasy in epistasis creates all of the above trends without requiring a biased distribution of epistasis. The idiosyncratic epistasis theory explains the universalities in mutational effects and evolutionary trajectories as emerging from randomness due to biological complexity.

Amino acid exchangeabilities vary across the tree of life.

Different amino acid pairs have drastically different relative exchangeabilities (REs), and accounting for this variation is an important and common practice in inferring phylogenies, testing selection, and predicting mutational effects, among other analyses. In all such endeavors, REs have been generally considered invariant among species; this assumption, however, has not been scrutinized. Using maximum likelihood to analyze 180 genome sequences, we estimated REs from 90 clades representing all three domains of life, and found numerous instances of substantial between-clade differences in REs. REs show more differences between orthologous proteins of different clades than unrelated proteins of the same clade, suggesting that REs are genome-wide, clade-specific features, probably a result of proteome-wide evolutionary changes in the physicochemical environments of amino acid residues. The discovery of among-clade RE variations cautions against assuming constant REs in various analyses and demonstrates a higher-than-expected complexity in mechanisms of proteome evolution.

Deciphering protein evolution and fitness landscapes with latent space models.

Protein sequences contain rich information about protein evolution, fitness landscapes, and stability. Here we investigate how latent space models trained using variational auto-encoders can infer these properties from sequences. Using both simulated and real sequences, we show that the low dimensional latent space representation of sequences, calculated using the encoder model, captures both evolutionary and ancestral relationships between sequences. Together with experimental fitness data and Gaussian process regression, the latent space representation also enables learning the protein fitness landscape in a continuous low dimensional space. Moreover, the model is also useful in predicting protein mutational stability landscapes and quantifying the importance of stability in shaping protein evolution. Overall, we illustrate that the latent space models learned using variational auto-encoders provide a mechanism for exploration of the rich data contained in protein sequences regarding evolution, fitness and stability and hence are well-suited to help guide protein engineering efforts.

Morphological and molecular convergences in mammalian phylogenetics.

Phylogenetic trees reconstructed from molecular sequences are often considered more reliable than those reconstructed from morphological characters, in part because convergent evolution, which confounds phylogenetic reconstruction, is believed to be rarer for molecular sequences than for morphologies. However, neither the validity of this belief nor its underlying cause is known. Here comparing thousands of characters of each type that have been used for inferring the phylogeny of mammals, we find that on average morphological characters indeed experience much more convergences than amino acid sites, but this disparity is explained by fewer states per character rather than an intrinsically higher susceptibility to convergence for morphologies than sequences. We show by computer simulation and actual data analysis that a simple method for identifying and removing convergence-prone characters improves phylogenetic accuracy, potentially enabling, when necessary, the inclusion of morphologies and hence fossils for reliable tree inference.

Y-Chromosome Structural Diversity in the Bonobo and Chimpanzee Lineages.

The male-specific regions of primate Y-chromosomes (MSY) are enriched for multi-copy genes highly expressed in the testis. These genes are located in large repetitive sequences arranged as palindromes, inverted-, and tandem repeats termed amplicons. In humans, these genes have critical roles in male fertility and are essential for the production of sperm. The structure of human and chimpanzee amplicon sequences show remarkable difference relative to the remainder of the genome, a difference that may be the result of intense selective pressure on male fertility. Four subspecies of common chimpanzees have undergone extended periods of isolation and appear to be in the early process of subspeciation. A recent study found amplicons enriched for testis-expressed genes on the primate X-chromosome the target of hard selective sweeps, and male-fertility genes on the Y-chromosome may also be the targets of selection. However, little is understood about Y-chromosome amplicon diversity within and across chimpanzee populations. Here, we analyze nine common chimpanzee (representing three subspecies: Pan troglodytes schweinfurthii, Pan troglodytes ellioti, and Pan troglodytes verus) and two bonobo (Pan paniscus) male whole-genome sequences to assess Y ampliconic copy-number diversity across the Pan genus. We observe that the copy number of Y chromosome amplicons is variable among chimpanzees and bonobos, and identify several lineage-specific patterns, including variable copy number of azoospermia candidates RBMY and DAZ We detect recurrent switchpoints of copy-number change along the ampliconic tracts across chimpanzee populations, which may be the result of localized genome instability or selective forces.

The development and application of a multiplex short tandem repeat (STR) system for identifying subspecies, individuals and sex in tigers.

Poaching and trans-boundary trafficking of tigers and body parts are threatening the world's last remaining wild tigers. Development of an efficient molecular genetic assay for tracing the origins of confiscated specimens will assist in law enforcement and wildlife forensics for this iconic flagship species. We developed a multiplex genotyping system "tigrisPlex" to simultaneously assess 22 short tandem repeat (STR, or microsatellite) loci and a gender-identifying SRY gene, all amplified in 4 reactions using as little as 1 ng of template DNA. With DNA samples used for between-run calibration, the system generates STR genotypes that are directly compatible with voucher tiger subspecies genetic profiles, hence making it possible to identify subspecies via bi-parentally inherited markers. We applied "tigrisPlex" to 12 confiscated specimens from Russia and identified 6 individuals (3 females and 3 males), each represented by duplicated samples and all designated as Amur tigers (Panthera tigris altaica) with high confidence. This STR multiplex system can serve as an effective and versatile approach for genetic profiling of both wild and captive tigers as well as confiscated tiger products, fulfilling various conservation needs for identifying the origins of tiger samples.

The genetic basis of white tigers.

The white tiger, an elusive Bengal tiger (Panthera tigris tigris) variant with white fur and dark stripes, has fascinated humans for centuries ever since its discovery in the jungles of India. Many white tigers in captivity are inbred in order to maintain this autosomal recessive trait and consequently suffer some health problems, leading to the controversial speculation that the white tiger mutation is perhaps a genetic defect. However, the genetic basis of this phenotype remains unknown. Here, we conducted genome-wide association mapping with restriction-site-associated DNA sequencing (RAD-seq) in a pedigree of 16 captive tigers segregating at the putative white locus, followed by whole-genome sequencing (WGS) of the three parents. Validation in 130 unrelated tigers identified the causative mutation to be an amino acid change (A477V) in the transporter protein SLC45A2. Three-dimensional homology modeling suggests that the substitution may partially block the transporter channel cavity and thus affect melanogenesis. We demonstrate the feasibility of combining RAD-seq and WGS to rapidly map exotic variants in nonmodel organisms. Our results identify the basis of the longstanding white tiger mystery as the same gene underlying color variation in human, horse, and chicken and highlight its significance as part of the species' natural polymorphism that is viable in the wild.