Joao S. Lopes

How host heterogeneity governs tuberculosis reinfection

Recurrent episodes of tuberculosis (TB) can be due to relapse of latent infection or exogenous reinfection, and discrimination is crucial for control planning. Molecular genotyping of Mycobacterium tuberculosis isolates offers concrete opportunities to measure the relative contribution of reinfection in recurrent disease. Here, a mathematical model of TB transmission is fitted to data from 14 molecular epidemiology studies, enabling the estimation of relevant epidemiological parameters. Meta-analysis reveals that rates of reinfection after successful treatment are higher than rates of new TB, raising an important question about the underlying mechanism. We formulate two alternative mechanisms within our model framework: (i) infection increases susceptibility to reinfection or (ii) infection affects individuals differentially, thereby recruiting high-risk individuals to the group at risk for reinfection. The second mechanism is better supported by the fittings to the data, suggesting that reinfection rates are inflated through a population phenomenon that occurs in the presence of heterogeneity in individual risk of infection. As a result, rates of reinfection are higher when measured at the population level even though they might be lower at the individual level. Finally, differential host recruitment is modulated by transmission intensity, being less pronounced when incidence is high.

Compositional biases among synonymous substitutions cause conflict between gene and protein trees for plastid origins

Archaeplastida (=Kingdom Plantae) are primary plastid-bearing organisms that evolved via the endosymbiotic association of a heterotrophic eukaryote host cell and a cyanobacterial endosymbiont approximately 1,400 Ma. Here, we present analyses of cyanobacterial and plastid genomes that show strongly conflicting phylogenies based on 75 plastid (or nuclear plastid-targeted) protein-coding genes and their direct translations to proteins. The conflict between genes and proteins is largely robust to the use of sophisticated data- and tree-heterogeneous composition models. However, by using nucleotide ambiguity codes to eliminate synonymous substitutions due to codon-degeneracy, we identify a composition bias, and dependent codon-usage bias, resulting from synonymous substitutions at all third codon positions and first codon positions of leucine and arginine, as the main cause for the conflicting phylogenetic signals. We argue that the protein-coding gene data analyses are likely misleading due to artifacts induced by convergent composition biases at first codon positions of leucine and arginine and at all third codon positions. Our analyses corroborate previous studies based on gene sequence analysis that suggest Cyanobacteria evolved by the early paraphyletic splitting of Gloeobacter and a specific Synechococcus strain (JA33Ab), with all other remaining cyanobacterial groups, including both unicellular and filamentous species, forming the sister-group to the Archaeplastida lineage. In addition, our analyses using better-fitting models suggest (but without statistically strong support) an early divergence of Glaucophyta within Archaeplastida, with the Rhodophyta (red algae), and Viridiplantae (green algae and land plants) forming a separate lineage.

Coestimation of recombination, substitution and molecular adaptation rates by approximate Bayesian computation

The estimation of parameters in molecular evolution may be biased when some processes are not considered. For example, the estimation of selection at the molecular level using codon-substitution models can have an upward bias when recombination is ignored. Here we address the joint estimation of recombination, molecular adaptation and substitution rates from coding sequences using approximate Bayesian computation (ABC). We describe the implementation of a regression-based strategy for choosing subsets of summary statistics for coding data, and show that this approach can accurately infer recombination allowing for intracodon recombination breakpoints, molecular adaptation and codon substitution rates. We demonstrate that our ABC approach can outperform other analytical methods under a variety of evolutionary scenarios. We also show that although the choice of the codon-substitution model is important, our inferences are robust to a moderate degree of model misspecification. In addition, we demonstrate that our approach can accurately choose the evolutionary model that best fits the data, providing an alternative for when the use of full-likelihood methods is impracticable. Finally, we applied our ABC method to co-estimate recombination, substitution and molecular adaptation rates from 24 published human immunodeficiency virus 1 coding data sets.

Unveiling time in dose-response models to infer host susceptibility to pathogens.

The biological effects of interventions to control infectious diseases typically depend on the intensity of pathogen challenge. As much as the levels of natural pathogen circulation vary over time and geographical location, the development of invariant efficacy measures is of major importance, even if only indirectly inferrable. Here a method is introduced to assess host susceptibility to pathogens, and applied to a detailed dataset generated by challenging groups of insect hosts (Drosophila melanogaster) with a range of pathogen (Drosophila C Virus) doses and recording survival over time. The experiment was replicated for flies carrying the Wolbachia symbiont, which is known to reduce host susceptibility to viral infections. The entire dataset is fitted by a novel quantitative framework that significantly extends classical methods for microbial risk assessment and provides accurate distributions of symbiont-induced protection. More generally, our data-driven modeling procedure provides novel insights for study design and analyses to assess interventions.

CodABC: A Computational Framework to Coestimate Recombination, Substitution, and Molecular Adaptation Rates by Approximate Bayesian Computation

The estimation of substitution and recombination rates can provide important insights into the molecular evolution of protein-coding sequences. Here, we present a new computational framework, called “CodABC,” to jointly estimate recombination, substitution and synonymous and nonsynonymous rates from coding data. CodABC uses approximate Bayesian computation with and without regression adjustment and implements a variety of codon models, intracodon recombination, and longitudinal sampling. CodABC can provide accurate joint parameter estimates from recombining coding sequences, often outperforming maximum-likelihood methods based on more approximate models. In addition, CodABC allows for the inclusion of several nuisance parameters such as those representing codon frequencies, transition matrices, heterogeneity across sites or invariable sites. CodABC is freely available from http://code.google.com/p/codabc/, includes a GUI, extensive documentation and ready-to-use examples, and can run in parallel on multicore machines.

SNP typing reveals similarity in Mycobacterium tuberculosis genetic diversity between Portugal and Northeast Brazil

Human tuberculosis is an infectious disease caused by bacteria from the Mycobacterium tuberculosis complex (MTBC). Although spoligotyping and MIRU-VNTR are standard methodologies in MTBC genetic epidemiology, recent studies suggest that Single Nucleotide Polymorphisms (SNP) are advantageous in phylogenetics and strain group/lineages identification. In this work we use a set of 79 SNPs to characterize 1987 MTBC isolates from Portugal and 141 from Northeast Brazil. All Brazilian samples were further characterized using spolygotyping. Phylogenetic analysis against a reference set revealed that about 95% of the isolates in both populations are singly attributed to bacterial lineage 4. Within this lineage, the most frequent strain groups in both Portugal and Brazil are LAM, followed by Haarlem and X. Contrary to these groups, strain group T showed a very different prevalence between Portugal (10%) and Brazil (1.5%). Spoligotype identification shows about 10% of mis-matches compared to the use of SNPs and a little more than 1% of strains unidentifiability. The mis-matches are observed in the most represented groups of our sample set (i.e., LAM and Haarlem) in almost the same proportion. Besides being more accurate in identifying strain groups/lineages, SNP-typing can also provide phylogenetic relationships between strain groups/lineages and, thus, indicate cases showing phylogenetic incongruence. Overall, the use of SNP-typing revealed striking similarities between MTBC populations from Portugal and Brazil.

Heterogeneity in symbiotic effects facilitates Wolbachia establishment in insect populations

Facultative vertically transmitted bacterial symbionts often manipulate its host’s reproductive biology and thus facilitate their persistence. Wolbachia is one such symbiont where frequency-dependent reproductive benefits are opposed by frequency-independent fitness costs leading to bistable dynamics. Introduction of carriers does not assure invasion unless the initial frequency is above a threshold determined by the balance of costs and benefits. Recent laboratory experiments have uncovered that Wolbachia also protects their hosts from pathogens. The expected consequence of this phenotype in natural environments is to lower the invasion threshold by a factor that increases with the extent of pathogen exposure. Here, we introduce a series of mathematical models to address how pathogen protection affects Wolbachia invasion. First, under homogeneous symbiotic effects, we obtain an analytical expression for the invasion threshold in terms of pathogen exposure, and find a regime where symbiont releases may result in elimination of the entire host population provided that abundance of virulent pathogens is high. Second, we distribute Wolbachia effects such that some carriers are totally protected and others not at all, and explore how this interplays with different pathogen intensities, to conclude that heterogeneity further lowers the threshold for Wolbachia invasion. Third, we replicate the analysis using a realistic distribution of protective effects and confirm that heterogeneity increases system resilience by reducing the odds of population collapse.

A theoretical framework to identify invariant thresholds in infectious disease epidemiology.

Setting global strategies and targets for disease prevention and control often involves mathematical models. Model structure is typically subject to intense scrutiny, such as confrontation with empirical data and alternative formulations, while a less frequently challenged aspect is the widely adopted reduction of parameters to their average values. Focusing on endemic diseases, we use a general transmission model to explain how mean field approximations decrease the estimated R0 from prevalence data, while threshold phenomena - such as the epidemic and reinfection thresholds - remain invariant. This results in an underestimation of the effort required to control disease, which may be particularly severe when the approximation inappropriately places transmission estimates below important thresholds. These concepts are widely applicable across endemic pathogen systems.

Impact of tuberculosis treatment length and adherence under different transmission intensities

Tuberculosis (TB) is a leading cause of human mortality due to infectious disease. Treatment default is a relevant factor which reduces therapeutic success and increases the risk of resistant TB. In this work we analyze the relation between treatment default and treatment length along with its consequence on the disease spreading. We use a stylized model structure to explore, systematically, the effects of varying treatment duration and compliance. We find that shortening treatment alone may not reduce TB prevalence, especially in regions where transmission intensity is high, indicating the necessity of complementing this action with increased compliance. A family of default functions relating the proportion of defaulters to the treatment length is considered and adjusted to a particular dataset. We find that the epidemiological benefits of shorter treatment regimens are tightly associated with increases in treatment compliance and depend on the epidemiological background.