Fisher's Geometric Model as a Tool to Study Speciation.

Interactions between alleles and across environments play an important role in the fitness of hybrids and are at the heart of the speciation process. Fitness landscapes capture these interactions and can be used to model hybrid fitness, helping us to interpret empirical observations and clarify verbal models. Here, we review recent progress in understanding hybridization outcomes through Fisher's geometric model, an intuitive and analytically tractable fitness landscape that captures many fitness patterns observed across taxa. We use case studies to show how the model parameters can be estimated from different types of data and discuss how these estimates can be used to make inferences about the divergence history and genetic architecture. We also highlight some areas where the model's predictions differ from alternative incompatibility-based models, such as the snowball effect and outlier patterns in genome scans.

Absence of Staphylococcus aureus in Wild Populations of Fish Supports a Spillover Hypothesis.

Staphylococcus aureus is a human commensal and opportunistic pathogen that also infects other animals. In humans and livestock, where S. aureus is most studied, strains are specialized for different host species. Recent studies have also found S. aureus in diverse wild animals. However, it remains unclear whether these isolates are also specialized for their hosts or whether their presence is due to repeated spillovers from source populations. This study focuses on S. aureus in fish, testing the spillover hypothesis in two ways. First, we examined 12 S. aureus isolates obtained from the internal and external organs of a farmed fish. While all isolates were from clonal complex 45, genomic diversity indicates repeated acquisition. The presence of a φSa3 prophage containing human immune evasion genes suggests that the source was originally human. Second, we tested for S. aureus in wild fish that were isolated from likely sources. In particular, we sampled 123 brown trout and their environment at 16 sites in the remote Scottish Highlands with variable levels of exposure to humans, birds, and livestock. This screen found no S. aureus infection in any of the wild populations or their environment. Together, these results support that the presence of S. aureus in fish and aquaculture is due to spillover from humans rather than specialization. Given the trends of increasing fish consumption, a better understanding of the dynamics of S. aureus spillover in aquaculture will mitigate future risks to fish and human health. IMPORTANCE Staphylococcus aureus is a human and livestock commensal but also an important pathogen responsible for high human mortality rates and economic losses in farming. Recent studies show that S. aureus is common in wild animals, including fish. However, we do not know whether these animals are part of the normal host range of S. aureus or whether infection is due to repeated spillover events from true S. aureus hosts. Answering this question has implications for public health and conservation. We find support for the spillover hypothesis by combining genome sequencing of S. aureus isolates from farmed fish and screens for S. aureus in isolated wild populations. The results imply that fish are unlikely to be a source of novel emergent S. aureus strains but highlight the prominence of the spillover of antibiotic-resistant bacteria from humans and livestock. This may affect both future fish disease potential and the risk of human food poisoning.

A 2-million-year-old ecosystem in Greenland uncovered by environmental DNA.

Late Pliocene and Early Pleistocene epochs 3.6 to 0.8 million years ago1 had climates resembling those forecasted under future warming2. Palaeoclimatic records show strong polar amplification with mean annual temperatures of 11-19 °C above contemporary values3,4. The biological communities inhabiting the Arctic during this time remain poorly known because fossils are rare5. Here we report an ancient environmental DNA6 (eDNA) record describing the rich plant and animal assemblages of the Kap København Formation in North Greenland, dated to around two million years ago. The record shows an open boreal forest ecosystem with mixed vegetation of poplar, birch and thuja trees, as well as a variety of Arctic and boreal shrubs and herbs, many of which had not previously been detected at the site from macrofossil and pollen records. The DNA record confirms the presence of hare and mitochondrial DNA from animals including mastodons, reindeer, rodents and geese, all ancestral to their present-day and late Pleistocene relatives. The presence of marine species including horseshoe crab and green algae support a warmer climate than today. The reconstructed ecosystem has no modern analogue. The survival of such ancient eDNA probably relates to its binding to mineral surfaces. Our findings open new areas of genetic research, demonstrating that it is possible to track the ecology and evolution of biological communities from two million years ago using ancient eDNA.

The diverse effects of phenotypic dominance on hybrid fitness.

When divergent populations interbreed, their alleles are brought together in hybrids. In the initial F1 cross, most divergent loci are heterozygous. Therefore, F1 fitness can be influenced by dominance effects that could not have been selected to function well together. We present a systematic study of these F1 dominance effects by introducing variable phenotypic dominance into Fisher's geometric model. We show that dominance often reduces hybrid fitness, which can generate optimal outbreeding followed by a steady decline in F1 fitness, as is often observed. We also show that "lucky" beneficial effects sometimes arise by chance, which might be important when hybrids can access novel environments. We then show that dominance can lead to violations of Haldane's Rule (reduced fitness of the heterogametic F1) but strengthens Darwin's Corollary (F1 fitness differences between cross directions). Taken together, results show that the effects of dominance on hybrid fitness can be surprisingly difficult to isolate, because they often resemble the effects of uniparental inheritance or expression. Nevertheless, we identify a pattern of environment-dependent heterosis that only dominance can explain, and for which there is some suggestive evidence. Our results also show how existing data set upper bounds on the size of dominance effects. These bounds could explain why additive models often provide good predictions for later-generation recombinant hybrids, even when dominance qualitatively changes outcomes for the F1.

A single introduction of wild rabbits triggered the biological invasion of Australia.

Biological invasions are a major cause of environmental and economic disruption. While ecological factors are key determinants of their success, the role of genetics has been more challenging to demonstrate. The colonization of Australia by the European rabbit is one of the most iconic and devastating biological invasions in recorded history. Here, we show that despite numerous introductions over a 70-y period, this invasion was triggered by a single release of a few animals that spread thousands of kilometers across the continent. We found genetic support for historical accounts that these were English rabbits imported in 1859 by a settler named Thomas Austin and traced the origin of the invasive population back to his birthplace in England. We also find evidence of additional introductions that established local populations but have not spread geographically. Combining genomic and historical data we show that, contrary to the earlier introductions, which consisted mostly of domestic animals, the invasive rabbits had wild ancestry. In New Zealand and Tasmania, rabbits also became a pest several decades after being introduced. We argue that the common denominator of these invasions was the arrival of a new genotype that was better adapted to the natural environment. These findings demonstrate how the genetic composition of invasive individuals can determine the success of an introduction and provide a mechanism by which multiple introductions can be required for a biological invasion.

Emergence of methicillin resistance predates the clinical use of antibiotics.

The discovery of antibiotics more than 80 years ago has led to considerable improvements in human and animal health. Although antibiotic resistance in environmental bacteria is ancient, resistance in human pathogens is thought to be a modern phenomenon that is driven by the clinical use of antibiotics1. Here we show that particular lineages of methicillin-resistant Staphylococcus aureus-a notorious human pathogen-appeared in European hedgehogs in the pre-antibiotic era. Subsequently, these lineages spread within the local hedgehog populations and between hedgehogs and secondary hosts, including livestock and humans. We also demonstrate that the hedgehog dermatophyte Trichophyton erinacei produces two ß-lactam antibiotics that provide a natural selective environment in which methicillin-resistant S. aureus isolates have an advantage over susceptible isolates. Together, these results suggest that methicillin resistance emerged in the pre-antibiotic era as a co-evolutionary adaptation of S. aureus to the colonization of dermatophyte-infected hedgehogs. The evolution of clinically relevant antibiotic-resistance genes in wild animals and the connectivity of natural, agricultural and human ecosystems demonstrate that the use of a One Health approach is critical for our understanding and management of antibiotic resistance, which is one of the biggest threats to global health, food security and development.

Prevalence and polymorphism of a mussel transmissible cancer in Europe.

Transmissible cancers are parasitic malignant cell lineages that have acquired the ability to infect new hosts from the same species, or sometimes related species. First described in dogs and Tasmanian devils, transmissible cancers were later discovered in some marine bivalves affected by a leukaemia-like disease. In Mytilus mussels, two lineages of bivalve transmissible neoplasia (BTN) have been described to date (MtrBTN1 and MtrBTN2), both of which emerged in a Mytilus trossulus founder individual. Here, we performed extensive screening of genetic chimerism, a hallmark of transmissible cancer, by genotyping 106 single nucleotide polymorphisms of 5,907 European Mytilus mussels. Genetic analysis allowed us to simultaneously obtain the genotype of hosts - Mytilus edulis, M. galloprovincialis or hybrids - and the genotype of tumours of heavily infected individuals. In addition, a subset of 222 individuals were systematically genotyped and analysed by histology to screen for possible nontransmissible cancers. We detected MtrBTN2 at low prevalence in M. edulis, and also in M. galloprovincialis and hybrids although at a much lower prevalence. No MtrBTN1 or new BTN were found, but eight individuals with nontransmissible neoplasia were observed at a single polluted site on the same sampling date. We observed a diversity of MtrBTN2 genotypes that appeared more introgressed or more ancestral than MtrBTN1 and reference healthy M. trossulus individuals. The observed polymorphism is probably due to somatic null alleles caused by structural variations or point mutations in primer-binding sites leading to enhanced detection of the host alleles. Despite low prevalence, two sublineages divergent by 10% fixed somatic null alleles and one nonsynonymous mtCOI (mitochondrial cytochrome oxidase I) substitution are cospreading in the same geographical area, suggesting a complex diversification of MtrBTN2 since its emergence and host species shift.

Large-scale genomic analysis of antimicrobial resistance in the zoonotic pathogen Streptococcus suis.

BACKGROUND: Antimicrobial resistance (AMR) is among the gravest threats to human health and food security worldwide. The use of antimicrobials in livestock production can lead to emergence of AMR, which can have direct effects on humans through spread of zoonotic disease. Pigs pose a particular risk as they are a source of zoonotic diseases and receive more antimicrobials than most other livestock. Here we use a large-scale genomic approach to characterise AMR in Streptococcus suis, a commensal found in most pigs, but which can also cause serious disease in both pigs and humans. RESULTS: We obtained replicated measures of Minimum Inhibitory Concentration (MIC) for 16 antibiotics, across a panel of 678 isolates, from the major pig-producing regions of the world. For several drugs, there was no natural separation into 'resistant' and 'susceptible', highlighting the need to treat MIC as a quantitative trait. We found differences in MICs between countries, consistent with their patterns of antimicrobial usage. AMR levels were high even for drugs not used to treat S. suis, with many multidrug-resistant isolates. Similar levels of resistance were found in pigs and humans from regions associated with zoonotic transmission. We next used whole genome sequences for each isolate to identify 43 candidate resistance determinants, 22 of which were novel in S. suis. The presence of these determinants explained most of the variation in MIC. But there were also interesting complications, including epistatic interactions, where known resistance alleles had no effect in some genetic backgrounds. Beta-lactam resistance involved many core genome variants of small effect, appearing in a characteristic order. CONCLUSIONS: We present a large dataset allowing the analysis of the multiple contributing factors to AMR in S. suis. The high levels of AMR in S. suis that we observe are reflected by antibiotic usage patterns but our results confirm the potential for genomic data to aid in the fight against AMR.

How do species barriers decay? Concordance and local introgression in mosaic hybrid zones of mussels.

The Mytilus complex of marine mussel species forms a mosaic of hybrid zones, found across temperate regions of the globe. This allows us to study 'replicated' instances of secondary contact between closely related species. Previous work on this complex has shown that local introgression is both widespread and highly heterogeneous, and has identified SNPs that are outliers of differentiation between lineages. Here, we developed an ancestry-informative panel of such SNPs. We then compared their frequencies in newly sampled populations, including samples from within the hybrid zones, and parental populations at different distances from the contact. Results show that close to the hybrid zones, some outlier loci are near to fixation for the heterospecific allele, suggesting enhanced local introgression, or the local sweep of a shared ancestral allele. Conversely, genomic cline analyses, treating local parental populations as the reference, reveal a globally high concordance among loci, albeit with a few signals of asymmetric introgression. Enhanced local introgression at specific loci is consistent with the early transfer of adaptive variants after contact, possibly including asymmetric bi-stable variants (Dobzhansky-Muller incompatibilities), or haplotypes loaded with fewer deleterious mutations. Having escaped one barrier, however, these variants can be trapped or delayed at the next barrier, confining the introgression locally. These results shed light on the decay of species barriers during phases of contact.

Life and Death of Selfish Genes: Comparative Genomics Reveals the Dynamic Evolution of Cytoplasmic Incompatibility.

Cytoplasmic incompatibility is a selfish reproductive manipulation induced by the endosymbiont Wolbachia in arthropods. In males Wolbachia modifies sperm, leading to embryonic mortality in crosses with Wolbachia-free females. In females, Wolbachia rescues the cross and allows development to proceed normally. This provides a reproductive advantage to infected females, allowing the maternally transmitted symbiont to spread rapidly through host populations. We identified homologs of the genes underlying this phenotype, cifA and cifB, in 52 of 71 new and published Wolbachia genome sequences. They are strongly associated with cytoplasmic incompatibility. There are up to seven copies of the genes in each genome, and phylogenetic analysis shows that Wolbachia frequently acquires new copies due to pervasive horizontal transfer between strains. In many cases, the genes have subsequently acquired loss-of-function mutations to become pseudogenes. As predicted by theory, this tends to occur first in cifB, whose sole function is to modify sperm, and then in cifA, which is required to rescue the cross in females. Although cif genes recombine, recombination is largely restricted to closely related homologs. This is predicted under a model of coevolution between sperm modification and embryonic rescue, where recombination between distantly related pairs of genes would create a self-incompatible strain. Together, these patterns of gene gain, loss, and recombination support evolutionary models of cytoplasmic incompatibility.

Genome Reduction Is Associated with Bacterial Pathogenicity across Different Scales of Temporal and Ecological Divergence.

Emerging bacterial pathogens threaten global health and food security, and so it is important to ask whether these transitions to pathogenicity have any common features. We present a systematic study of the claim that pathogenicity is associated with genome reduction and gene loss. We compare broad-scale patterns across all bacteria, with detailed analyses of Streptococcus suis, an emerging zoonotic pathogen of pigs, which has undergone multiple transitions between disease and carriage forms. We find that pathogenicity is consistently associated with reduced genome size across three scales of divergence (between species within genera, and between and within genetic clusters of S. suis). Although genome reduction is also found in mutualist and commensal bacterial endosymbionts, genome reduction in pathogens cannot be solely attributed to the features of their ecology that they share with these species, that is, host restriction or intracellularity. Moreover, other typical correlates of genome reduction in endosymbionts (reduced metabolic capacity, reduced GC content, and the transient expansion of nonfunctional elements) are not consistently observed in pathogens. Together, our results indicate that genome reduction is a consistent correlate of pathogenicity in bacteria.

The geometry and genetics of hybridization.

When divergent populations form hybrids, hybrid fitness can vary with genome composition, current environmental conditions, and the divergence history of the populations. We develop analytical predictions for hybrid fitness, which incorporate all three factors. The predictions are based on Fisher's geometric model, and apply to a wide range of population genetic parameter regimes and divergence conditions, including allopatry and parapatry, local adaptation, and drift. Results show that hybrid fitness can be decomposed into intrinsic effects of admixture and heterozygosity, and extrinsic effects of the (local) adaptedness of the parental lines. Effect sizes are determined by a handful of geometric distances, which have a simple biological interpretation. These distances also reflect the mode and amount of divergence, such that there is convergence toward a characteristic pattern of intrinsic isolation. We next connect our results to the quantitative genetics of line crosses in variable or patchy environments. This means that the geometrical distances can be estimated from cross data, and provides a simple interpretation of the "composite effects." Finally, we develop extensions to the model, involving selectively induced disequilibria, and variable phenotypic dominance. The geometry of fitness landscapes provides a unifying framework for understanding speciation, and wider patterns of hybrid fitness.

Replicated anthropogenic hybridisations reveal parallel patterns of admixture in marine mussels.

Human-mediated transport creates secondary contacts between genetically differentiated lineages, bringing new opportunities for gene exchange. When similar introductions occur in different places, they provide informally replicated experiments for studying hybridisation. We here examined 4,279 Mytilus mussels, sampled in Europe and genotyped with 77 ancestry-informative markers. We identified a type of introduced mussels, called "dock mussels," associated with port habitats and displaying a particular genetic signal of admixture between M. edulis and the Mediterranean lineage of M. galloprovincialis. These mussels exhibit similarities in their ancestry compositions, regardless of the local native genetic backgrounds and the distance separating colonised ports. We observed fine-scale genetic shifts at the port entrance, at scales below natural dispersal distance. Such sharp clines do not fit with migration-selection tension zone models, and instead suggest habitat choice and early-stage adaptation to the port environment, possibly coupled with connectivity barriers. Variations in the spread and admixture patterns of dock mussels seem to be influenced by the local native genetic backgrounds encountered. We next examined departures from the average admixture rate at different loci, and compared human-mediated admixture events, to naturally admixed populations and experimental crosses. When the same M. galloprovincialis background was involved, positive correlations in the departures of loci across locations were found; but when different backgrounds were involved, no or negative correlations were observed. While some observed positive correlations might be best explained by a shared history and saltatory colonisation, others are likely produced by parallel selective events. Altogether, genome-wide effect of admixture seems repeatable and more dependent on genetic background than environmental context. Our results pave the way towards further genomic analyses of admixture, and monitoring of the spread of dock mussels both at large and at fine spacial scales.

Insights into herpesvirus assembly from the structure of the pUL7:pUL51 complex.

Herpesviruses acquire their membrane envelopes in the cytoplasm of infected cells via a molecular mechanism that remains unclear. Herpes simplex virus (HSV)-1 proteins pUL7 and pUL51 form a complex required for efficient virus envelopment. We show that interaction between homologues of pUL7 and pUL51 is conserved across human herpesviruses, as is their association with trans-Golgi membranes. We characterized the HSV-1 pUL7:pUL51 complex by solution scattering and chemical crosslinking, revealing a 1:2 complex that can form higher-order oligomers in solution, and we solved the crystal structure of the core pUL7:pUL51 heterodimer. While pUL7 adopts a previously-unseen compact fold, the helix-turn-helix conformation of pUL51 resembles the cellular endosomal complex required for transport (ESCRT)-III component CHMP4B and pUL51 forms ESCRT-III-like filaments, suggesting a direct role for pUL51 in promoting membrane scission during virus assembly. Our results provide a structural framework for understanding the role of the conserved pUL7:pUL51 complex in herpesvirus assembly.

Most people suffer from occasional cold sores, which are caused by the herpes simplex virus. This virus causes infections that last your entire life, but for the most part it lies dormant in your cells and reactivates only at times of stress. When it reactivates, the virus manipulates host cells to make new virus particles that may spread the infection to other people. Like many other viruses, herpes simplex viruses also steal jelly-like structures known as membranes from their host cells to form protective coats around new virus particles. In cells from humans and other animals, proteins belonging to a molecular machine known as ESCRT form filaments that bend and break membranes as the cells require. Many viruses hijack the ESCRT machinery to wrap membranes around new virus particles. However, herpes simplex viruses do not follow the usual rules for activating this machine. Instead, they rely on two viral proteins called pUL7 and pUL51 to hot-wire the ESCRT machinery. Previous studies have shown that these two proteins bind to each other, but it remained unclear how they work. Butt et al. used a combination of biochemical and biophysical techniques to solve the three-dimensional structures of pUL7 and pUL51 when bound to each other. The experiments determined that the structure of pUL51 resembles the structures of different components in the ESCRT machinery. Like the ESCRT proteins, pUL51 formed filaments, suggesting that pUL51 bends membranes in cells and that pUL7 blocks it from doing so until the time is right. Further experiments showed that the equivalents of pUL7 and pUL51 in other members of the herpes virus family also bind to each other in a similar way. These findings reveal that herpes simplex viruses and their close relatives have evolved a different strategy than many other viruses to steal membranes from host cells. Interfering with this mechanism may provide new avenues for designing drugs or improving vaccines against these viruses. The pUL7 and pUL51 proteins may also inspire new tools in biotechnology that could precisely control the shapes of biological membranes.

Ice-Age Climate Adaptations Trap the Alpine Marmot in a State of Low Genetic Diversity.

Some species responded successfully to prehistoric changes in climate [1, 2], while others failed to adapt and became extinct [3]. The factors that determine successful climate adaptation remain poorly understood. We constructed a reference genome and studied physiological adaptations in the Alpine marmot (Marmota marmota), a large ground-dwelling squirrel exquisitely adapted to the "ice-age" climate of the Pleistocene steppe [4, 5]. Since the disappearance of this habitat, the rodent persists in large numbers in the high-altitude Alpine meadow [6, 7]. Genome and metabolome showed evidence of adaptation consistent with cold climate, affecting white adipose tissue. Conversely, however, we found that the Alpine marmot has levels of genetic variation that are among the lowest for mammals, such that deleterious mutations are less effectively purged. Our data rule out typical explanations for low diversity, such as high levels of consanguineous mating, or a very recent bottleneck. Instead, ancient demographic reconstruction revealed that genetic diversity was lost during the climate shifts of the Pleistocene and has not recovered, despite the current high population size. We attribute this slow recovery to the marmot's adaptive life history. The case of the Alpine marmot reveals a complicated relationship between climatic changes, genetic diversity, and conservation status. It shows that species of extremely low genetic diversity can be very successful and persist over thousands of years, but also that climate-adapted life history can trap a species in a persistent state of low genetic diversity.

Advancing Cancer Research in Africa Through Early-Career Awards: The BIG Cat Initiative.

PURPOSE: The burden of cancer in Africa is growing rapidly, and increased cancer research on the continent is a critical component of an effective response. In 2010, the US National Cancer Institute, in partnership with the African Organization for Research and Training in Cancer, launched the Beginning Investigator Grant for Catalytic Research (BIG Cat) initiative to support cancer research projects conducted by early-career African investigators. METHODS: To date, BIG Cat has provided 18 awards of up to $50,000 to support 2-year cancer research projects. In 2017, the National Cancer Institute evaluated BIG Cat's early outcomes for cancer research and impacts on career development and local cancer research capacity. Data collection consisted of a review of project documentation and a survey fielded to the 12 investigators who had completed their BIG Cat awards. RESULTS: BIG Cat-supported research projects have generated locally relevant findings that address a range of cancer sites and multiple areas of scientific interest. The 11 survey respondents produced 43 scholarly products (e.g., publications, presentations) about findings from their BIG Cat research. They reported increases in cancer research funding applications and awards after receipt of the BIG Cat award compared with before the award. They also reported increased resources for cancer research, participation in teaching and mentoring on cancer research, and supervision of cancer research staff. Investigators identified scientific mentoring as a key facilitator of the success of their BIG Cat projects and limited time and funding as key challenges. CONCLUSION: Findings provide early evidence that BIG Cat advanced locally relevant cancer research and facilitated career advancement and development of local cancer research capacity. Findings have implications for the design of future related efforts.

The distribution of epistasis on simple fitness landscapes.

Fitness interactions between mutations can influence a population's evolution in many different ways. While epistatic effects are difficult to measure precisely, important information is captured by the mean and variance of log fitnesses for individuals carrying different numbers of mutations. We derive predictions for these quantities from a class of simple fitness landscapes, based on models of optimizing selection on quantitative traits. We also explore extensions to the models, including modular pleiotropy, variable effect sizes, mutational bias and maladaptation of the wild type. We illustrate our approach by reanalysing a large dataset of mutant effects in a yeast snoRNA (small nucleolar RNA). Though characterized by some large epistatic effects, these data give a good overall fit to the non-epistatic null model, suggesting that epistasis might have limited influence on the evolutionary dynamics in this system. We also show how the amount of epistasis depends on both the underlying fitness landscape and the distribution of mutations, and so is expected to vary in consistent ways between new mutations, standing variation and fixed mutations.

Coadapted genomes and selection on hybrids: Fisher's geometric model explains a variety of empirical patterns.

Natural selection plays a variety of roles in hybridization, speciation, and admixture. Most research has focused on two extreme cases: crosses between closely related inbred lines, where hybrids are fitter than their parents, or crosses between effectively isolated species, where hybrids suffer severe breakdown. But many natural populations must fall into intermediate regimes, with multiple types of gene interaction, and these are more difficult to study. Here, we develop a simple fitness landscape model, and show that it naturally interpolates between previous modeling approaches, which were designed for the extreme cases, and invoke either mildly deleterious recessives, or discrete hybrid incompatibilities. Our model yields several new predictions, which we test with genomic data from Mytilus mussels, and published data from plants (Zea, Populus, and Senecio) and animals (Mus, Teleogryllus, and Drosophila). The predictions are generally supported, and the model explains a number of surprising empirical patterns. Our approach enables novel and complementary uses of genome-wide datasets, which do not depend on identifying outlier loci, or "speciation genes" with anomalous effects. Given its simplicity and flexibility, and its predictive successes with a wide range of data, the approach should be readily extendable to other outstanding questions in the study of hybridization.

The divergence history of European blue mussel species reconstructed from Approximate Bayesian Computation: the effects of sequencing techniques and sampling strategies.

Genome-scale diversity data are increasingly available in a variety of biological systems, and can be used to reconstruct the past evolutionary history of species divergence. However, extracting the full demographic information from these data is not trivial, and requires inferential methods that account for the diversity of coalescent histories throughout the genome. Here, we evaluate the potential and limitations of one such approach. We reexamine a well-known system of mussel sister species, using the joint site frequency spectrum (jSFS) of synonymous mutations computed either from exome capture or RNA-seq, in an Approximate Bayesian Computation (ABC) framework. We first assess the best sampling strategy (number of: individuals, loci, and bins in the jSFS), and show that model selection is robust to variation in the number of individuals and loci. In contrast, different binning choices when summarizing the jSFS, strongly affect the results: including classes of low and high frequency shared polymorphisms can more effectively reveal recent migration events. We then take advantage of the flexibility of ABC to compare more realistic models of speciation, including variation in migration rates through time (i.e., periodic connectivity) and across genes (i.e., genome-wide heterogeneity in migration rates). We show that these models were consistently selected as the most probable, suggesting that mussels have experienced a complex history of gene flow during divergence and that the species boundary is semi-permeable. Our work provides a comprehensive evaluation of ABC demographic inference in mussels based on the coding jSFS, and supplies guidelines for employing different sequencing techniques and sampling strategies. We emphasize, perhaps surprisingly, that inferences are less limited by the volume of data, than by the way in which they are analyzed.