Search bioRxiv⌕ Search
Home/Archive/bioRxiv

bioRxiv · 10.1101/085993

Estimation of sub-epidemic dynamics by means of Sequential Monte Carlo Approximate Bayesian Computation: an application to the Swiss HIV Cohort Study

Ibeh, N.·Aris-Brosou, S.
View original source ↗Export metadata ↓

Abstract

Our ability to accurately infer transmission patterns of infectious diseases is critical to monitor both their spread and the efficacy of public health policies. The use of phylogenetic methods for the reconstruction of viral ancestral relationships has garnered increasing interest, particularly in the characterization of HIV epidemics and sub-epidemics. In the case of this virus, the Swiss HIV Cohort Study (SHCS) contains a wide breadth of genomic data that have been widely used as a means of applying such methods. However, current approaches for quantifying the epidemiological dynamics of diseases are computationally intensive, and fail to scale well with this magnitude of data. To address this issue, we re-implement an Approximate Bayesian Computation (ABC) approach based on sequential Monte Carlo (SMC). By means of simulations, we demonstrate that our implementation is capable of inferring key epidemiological parameters of the Swiss HIV epidemic accurately, and that sampling intensity has no significant effect on the accuracy of our estimates. Applied to a subset of HIV sequences from the SHCS, we show that we can distinguish sub-epidemics that are circulating in culturally distinct Swiss regions. Given these findings, we propose that ABC-SMC samplers will allow us to evaluate the impact of new public health policies, such as the implementation of a needle exchange program in the case of HIV, based on genetic data sampled before and after the implementation of a new policy.

Explore related subjects

evolutionary biologyPreprint

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Ibeh, N., Aris-Brosou, S.. 2016-11-07. Estimation of sub-epidemic dynamics by means of Sequential Monte Carlo Approximate Bayesian Computation: an application to the Swiss HIV Cohort Study. https://doi.org/10.1101/085993

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related preprints

bioRxiv2026

Integrative taxonomy of the North American pink flat-backed millipedes (Polydesmidae: Pseudopolydesmus Attems, 1898)

Millipedes of the genus Pseudopolydesmus Attems, 1898 are common millipedes in North America. They are the largest, in terms of body size, and most geographically wide-ranging of the family Polydesmidae, and are ubiquitous in eastern and central North American forests. Eight species have been named based on the morphology of the genitalia, but these have not been investigated genetically. Here, we take an integrated taxonomic approach, using five genes to infer a molecular phylogeny of Pseudopolydesmus species, and analyze a dataset of 15 morphological characters in an effort to delimit species in the genus. 1,148 Pseudopolydesmus specimens were newly collected for this study, and all known species were sampled. These specimens were used for analysis, and were combined with museum specimens and literature records into a database of 3,462 unique specimen records. This database was used to produce distribution maps for each species in the genus. The maximum likelihood phylogeny found support for the eight species in the genus, supporting previous morphological species hypotheses. However, previous morphology-based hypotheses regarding evolutionary relationships between the species were not supported by the analysis. The morphological analysis, performed with a modified PCA, did not delimit species clusters as well as the genitalic and genetic differences did, due to comparatively uniform somatic morphology in the genus. However, our results show that genitalic morphology, combined with genetic analysis, provide congruent species delimitation hypotheses for Pseudopolydesmus, and these lines of evidence may prove useful in delimiting species and generic boundaries in other Polydesmidae worldwide.

evolutionary biology↗
bioRxiv2026

The pupal stage shapes life-history trade-offs across metamorphosis

Dealing with ecological stressors during development often comes at the expense of later-life performance. Although such trade-offs are well studied as two-stage systems such as larva-adult trade-offs, they are rarely examined across multiple developmental stages, limiting our understanding of how ecological effects are integrated throughout complex life cycles. We addressed this gap using a multistage experimental framework in Aedes aegypti, where we exposed individuals to predation risk during the larval stage, the pupal stage, both stages, or neither, and quantified developmental, physiological, and adult fitness traits. Contrary to expectations, predation risk across successive stages did not produce cumulative fitness costs, and traits varied in their responses. Both males and females emerged faster when exposed to risk, with lighter bodies, but males paid an additional cost of reduced starvation resistance when the pupal stage was exposed. Surprisingly, despite their lower body mass, females showed enhanced reproductive performance when both stages were exposed to risk. Overall, responses were stage- and sex-specific, with exposure during the non-feeding pupal stage imposing greater life-history costs than larval exposure alone. Measurements across the larva-pupa-adult continuum further revealed that traits such as body mass and energy reserves respond differently before and after metamorphosis, highlighting the role of the pupa in mediating trade-offs. Our findings identify the pupal stage as a critical mediator of life-history trade-offs and demonstrate the value of a multistage framework for understanding the role of each stage in life-history evolution.

evolutionary biology↗
bioRxiv2026

Cell-lineage selection interacts with development and life history to shape the rates and identity of heritable mutations

Mutation rates determine the supply of heritable variation and are central to evolutionary theory, yet they are usually treated as properties of molecular mutational processes. We hypothesised that mutation rates are also shaped by selection among cell lineages within organisms and that the strength of this process depends on developmental and life-history traits. To test the extent of this idea, we combined mathematical modelling with experiments in Drosophila melanogaster and treated mutation rates as emergent properties of interactions among molecular mutagenesis, development, life history, and cell-lineage selection. The model predicts that even modest effects of mutations on cellular fitness can substantially alter the probability that mutations are transmitted, and that these effects depend strongly on developmental architecture, germline dynamics, and time to reproduction. In Drosophila, we experimentally demonstrate biased transmission of a mutation affecting cellular proliferation and show that this bias changes over the reproductive lifespan of the individual. Extending the model genome-wide reveals that the cellular distribution of fitness effects and its interaction with developmental parameters influence both the estimated mutation rate and the composition of transmitted mutations, producing transmission biases that favour mutations with particular cellular fitness effects. These results suggest that developmental and life-history parameters determine how cellular fitness effects translate into transmission bias; therefore, estimates of heritable mutational input can vary from species to species even under constant molecular mutagenesis. By filtering mutations before they enter populations, developmental and life-history evolution can modify both the rate and the nature of heritable variation and thereby influence evolutionary trajectories.

evolutionary biology↗