Deep Learning and Likelihood Approaches for Viral Phylogeography Converge on the Same Answers Whether the Inference Model Is Right or Wrong.

Analysis of phylogenetic trees has become an essential tool in epidemiology. Likelihood-based methods fit models to phylogenies to draw inferences about the phylodynamics and history of viral transmission. However, these methods are often computationally expensive, which limits the complexity and realism of phylodynamic models and makes them ill-suited for informing policy decisions in real-time during rapidly developing outbreaks. Likelihood-free methods using deep learning are pushing the boundaries of inference beyond these constraints. In this paper, we extend, compare, and contrast a recently developed deep learning method for likelihood-free inference from trees. We trained multiple deep neural networks using phylogenies from simulated outbreaks that spread among 5 locations and found they achieve close to the same levels of accuracy as Bayesian inference under the true simulation model. We compared robustness to model misspecification of a trained neural network to that of a Bayesian method. We found that both models had comparable performance, converging on similar biases. We also implemented a method of uncertainty quantification called conformalized quantile regression that we demonstrate has similar patterns of sensitivity to model misspecification as Bayesian highest posterior density (HPD) and greatly overlap with HPDs, but have lower precision (more conservative). Finally, we trained and tested a neural network against phylogeographic data from a recent study of the SARS-Cov-2 pandemic in Europe and obtained similar estimates of region-specific epidemiological parameters and the location of the common ancestor in Europe. Along with being as accurate and robust as likelihood-based methods, our trained neural networks are on average over 3 orders of magnitude faster after training. Our results support the notion that neural networks can be trained with simulated data to accurately mimic the good and bad statistical properties of the likelihood functions of generative phylogenetic models.

Sexual dimorphism in chimpanzee (Pan troglodytes schweinfurthii) and human age-specific fertility.

Across vertebrates, species with intense male mating competition and high levels of sexual dimorphism in body size generally exhibit dimorphism in age-specific fertility. Compared with females, males show later ages at first reproduction and earlier reproductive senescence because they take longer to attain adult body size and musculature, and maintain peak condition for a limited time. This normally yields a shorter male duration of effective breeding, but this reduction might be attenuated in species that frequently use coalitionary aggression. Here, we present comparative genetic and demographic data on chimpanzees from three long-term study communities (Kanyawara: Kibale National Park, Uganda; Mitumba and Kasekela: Gombe National Park, Tanzania), comprising 581 male risk years and 112 infants, to characterize male age-specific fertility. For comparison, we update estimates from female chimpanzees in the same sites and append a sample of human foragers (the Tanzanian Hadza). Consistent with the idea that aggressive mating competition favors youth, chimpanzee males attained a higher maximum fertility than females, followed by a steeper decline with age. Males did not show a delay in reproduction compared with females, however, as adolescents in both sites successfully reproduced by targeting young, subfecund females, who were less attractive to adults. Gombe males showed earlier reproductive senescence and a shorter duration of effective breeding than Gombe females. By contrast, older males in Kanyawara generally continued to reproduce, apparently by forming coalitions with the alpha. Hadza foragers showed a distinct pattern of sexual dimorphism in age-specific fertility as, compared with women, men gained conceptions later but continued reproducing longer. In sum, both humans and chimpanzees showed sexual dimorphism in age-specific fertility that deviated from predictions drawn from primates with more extreme body size dimorphism, suggesting altered dynamics of male-male competition in the two lineages. In both species, coalitions appear important for extending male reproductive careers.

Generation of an immortalized erythroid progenitor cell line from peripheral blood: A model system for the functional analysis of Plasmodium spp. invasion.

Malaria pathogenesis is caused by the replication of Plasmodium parasites within the red blood cells (RBCs) of the vertebrate host. This selective pressure has favored the evolution of protective polymorphisms in erythrocyte proteins, a subset of which serve as cognate receptors for parasite invasion ligands. Recently, the generation of RBCs from immortalized hematopoietic stem cells (HSCs) has offered a more tractable system for genetic manipulation and long-term in vitro culture, enabling elucidation of the functional determinants of host susceptibility in vitro. Here we report the generation of an immortalized erythroid progenitor cell line (EJ cells) from as few as 100 000 peripheral blood mononuclear cells. It offers a robust method for the creation of customized model systems from small volumes of peripheral blood. The EJ cell differentiation mirrored erythropoiesis of primary HSCs, yielding orthochromatic erythroblasts and enucleated RBCs after eight days (ejRBCs). The ejRBCs supported invasion by both P. vivax and P. falciparum. To demonstrate the genetic tractability of this system, we used CRISPR/Cas9 to disrupt the Duffy Antigen/Receptor for Chemokines (DARC) gene, which encodes the canonical receptor of P. vivax in humans. Invasion of P. vivax into this DARC-knockout cell line was strongly inhibited providing direct genetic evidence that P. vivax requires DARC for RBC invasion. Further, genetic complementation of DARC restored P. vivax invasion. Taken together, the peripheral blood immortalization method presented here offers the capacity to generate biologically representative model systems for studies of blood-stage malaria invasion from the peripheral blood of donors harboring unique genetic backgrounds, or rare polymorphisms.

Lethal Respiratory Disease Associated with Human Rhinovirus C in Wild Chimpanzees, Uganda, 2013.

We describe a lethal respiratory outbreak among wild chimpanzees in Uganda in 2013 for which molecular and epidemiologic analyses implicate human rhinovirus C as the cause. Postmortem samples from an infant chimpanzee yielded near-complete genome sequences throughout the respiratory tract; other pathogens were absent. Epidemiologic modeling estimated the basic reproductive number (R0) for the epidemic as 1.83, consistent with the common cold in humans. Genotyping of 41 chimpanzees and examination of 24 published chimpanzee genomes from subspecies across Africa showed universal homozygosity for the cadherin-related family member 3 CDHR3-Y529 allele, which increases risk for rhinovirus C infection and asthma in human children. These results indicate that chimpanzees exhibit a species-wide genetic susceptibility to rhinovirus C and that this virus, heretofore considered a uniquely human pathogen, can cross primate species barriers and threatens wild apes. We advocate engineering interventions and prevention strategies for rhinovirus infections for both humans and wild apes.

The ecology and epidemiology of malaria parasitism in wild chimpanzee reservoirs.

Chimpanzees (Pan troglodytes) harbor rich assemblages of malaria parasites, including three species closely related to P. falciparum (sub-genus Laverania), the most malignant human malaria parasite. Here, we characterize the ecology and epidemiology of malaria infection in wild chimpanzee reservoirs. We used molecular assays to screen chimpanzee fecal samples, collected longitudinally and cross-sectionally from wild populations, for malaria parasite mitochondrial DNA. We found that chimpanzee malaria parasitism has an early age of onset and varies seasonally in prevalence. A subset of samples revealed Hepatocystis mitochondrial DNA, with phylogenetic analyses suggesting that Hepatocystis appears to cross species barriers more easily than Laverania. Longitudinal and cross-sectional sampling independently support the hypothesis that mean ambient temperature drives spatiotemporal variation in chimpanzee Laverania infection. Infection probability peaked at ~24.5 °C, consistent with the empirical transmission optimum of P. falciparum in humans. Forest cover was also positively correlated with spatial variation in Laverania prevalence, consistent with the observation that forest-dwelling Anophelines are the primary vectors. Extrapolating these relationships across equatorial Africa, we map spatiotemporal variation in the suitability of chimpanzee habitat for Laverania transmission, offering a hypothetical baseline indicator of human exposure risk.

Simultaneous outbreaks of respiratory disease in wild chimpanzees caused by distinct viruses of human origin.

Respiratory viruses of human origin infect wild apes across Africa, sometimes lethally. Here we report simultaneous outbreaks of two distinct human respiratory viruses, human metapneumovirus (MPV; Pneumoviridae: Metapneumovirus) and human respirovirus 3 (HRV3; Paramyxoviridae; Respirovirus, formerly known as parainfluenza virus 3), in two chimpanzee (Pan troglodytes schweinfurthii) communities in the same forest in Uganda in December 2016 and January 2017. The viruses were absent before the outbreaks, but each was present in ill chimpanzees from one community during the outbreak period. Clinical signs and gross pathologic changes in affected chimpanzees closely mirrored symptoms and pathology commonly observed in humans for each virus. Epidemiologic modelling showed that MPV and HRV3 were similarly transmissible (R0 of 1.27 and 1.48, respectively), but MPV caused 12.2% mortality mainly in infants and older chimpanzees, whereas HRV3 caused no direct mortality. These results are consistent with the higher virulence of MPV than HRV3 in humans, although both MPV and HRV3 cause a significant global disease burden. Both viruses clustered phylogenetically within groups of known human variants, with MPV closely related to a lethal 2009 variant from mountain gorillas (Gorilla beringei beringei), suggesting two independent and simultaneous reverse zoonotic origins, either directly from humans or via intermediary hosts. These findings expand our knowledge of human origin viruses threatening wild chimpanzees and suggest that such viruses might be differentiated by their comparative epidemiological dynamics and pathogenicity in wild apes. Our results also caution against assuming common causation in coincident outbreaks.

Risk factors for respiratory illness in a community of wild chimpanzees (Pan troglodytes schweinfurthii).

Respiratory illnesses have caused significant mortality in African great ape populations. While much effort has been given to identifying the responsible pathogens, little is known about the factors that influence disease transmission or individual susceptibility. In the Kanyawara community of wild chimpanzees, respiratory illness has been the leading cause of mortality over 31 years, contributing to 27% of deaths. Deaths were common in all age groups except juveniles. Over 22 years of health observations, respiratory signs were rare among infants and most common among older adults of both sexes. Respiratory signs were also common among males during the transition to adulthood (ages 10-20 years), particularly among those of low rank. Respiratory signs peaked conspicuously in March, a pattern that we could not explain after modelling climatic factors, group sizes, diet or exposure to humans. Furthermore, rates of respiratory illness in the chimpanzees did not track seasonal rates of illness in the nearby village. Our data indicate that the epidemiology of chimpanzee respiratory illness warrants more investigation but clearly differs in important ways from humans. Findings on individual susceptibility patterns suggest that respiratory signs are a robust indicator for investigating immunocompetence in wild chimpanzees.

Molecular interactions governing host-specificity of blood stage malaria parasites.

Non-human primates harbor diverse species of malaria parasites, including the progenitors of Plasmodium falciparum and Plasmodium vivax. Cross-species transmission of some malaria parasites-most notably the macaque parasite, Plasmodium knowlesi-continues to this day, compelling the scientific community to ask whether these zoonoses could impede malaria control efforts by acting as a source of recurrent human infection. Host-restriction varies considerably among parasite species and is governed by both ecological and molecular variables. In particular, the efficiency of red blood cell invasion constitutes a prominent barrier to zoonotic emergence. Although proteins expressed upon the erythrocyte surface exhibit considerable diversity both within and among hosts, malaria parasites have adapted to this heterogeneity via the expansion of protein families associated with invasion, offering redundant mechanisms of host cell entry. This molecular toolkit may enable some parasites to circumvent host barriers, potentially yielding host shifts upon subsequent adaptation. Recent studies have begun to elucidate the molecular determinants of host-specificity, as well as the mechanisms that malaria parasites use to overcome these restrictions. We review recent studies concerning host tropism in the context of erythrocyte invasion by focusing on three malaria parasites that span the zoonotic spectrum: P. falciparum, P. knowlesi, and P. vivax.

Wild bonobos host geographically restricted malaria parasites including a putative new Laverania species.

Malaria parasites, though widespread among wild chimpanzees and gorillas, have not been detected in bonobos. Here, we show that wild-living bonobos are endemically Plasmodium infected in the eastern-most part of their range. Testing 1556 faecal samples from 11 field sites, we identify high prevalence Laverania infections in the Tshuapa-Lomami-Lualaba (TL2) area, but not at other locations across the Congo. TL2 bonobos harbour P. gaboni, formerly only found in chimpanzees, as well as a potential new species, Plasmodium lomamiensis sp. nov. Rare co-infections with non-Laverania parasites were also observed. Phylogenetic relationships among Laverania species are consistent with co-divergence with their gorilla, chimpanzee and bonobo hosts, suggesting a timescale for their evolution. The absence of Plasmodium from most field sites could not be explained by parasite seasonality, nor by bonobo population structure, diet or gut microbiota. Thus, the geographic restriction of bonobo Plasmodium reflects still unidentified factors that likely influence parasite transmission.