DENcode: A model for haplotype-informed transmission probability of dengue virus

Area of Science:

• * Epidemiology and Public Health
• * Viral Genomics and Phylogenetics
• * Mathematical Modeling of Infectious Diseases

Background:

• * Understanding dengue virus transmission is crucial for effective control, yet current methods often provide incomplete networks due to sampling limitations and unobserved transmission routes.
• * Phylogenetic methods reveal evolutionary relationships but do not explicitly define probabilistic transmission events between individual cases.
• * Accurate reconstruction of dengue transmission networks requires integrating diverse data sources to capture complex epidemiological dynamics.

Purpose of the Study:

• * To develop and validate DENcode, a novel computational framework for estimating the likelihood of vector-mediated dengue virus transmission between individual infections.
• * To integrate epidemiological parameters (extrinsic incubation period, human infectiousness) with pathogen genetic similarity (haplotype/consensus sequence divergence) for improved transmission inference.
• * To construct and analyze dengue transmission networks using pairwise linkage probabilities derived from the DENcode framework.

Main Methods:

• * Development of DENcode, combining a temperature- and time-modulated epidemiological kernel with a phylogenetically informed genetic similarity kernel.
• * Utilizing patristic distances derived from viral haplotypes or consensus sequences to quantify genetic relatedness between dengue virus samples.
• * Validation using a dataset of 90 dengue infections from Colombo, Sri Lanka (2017-2020), employing Monte Carlo simulations and sensitivity analyses (ablation experiments).

Main Results:

• * DENcode provided stable estimates with narrow credible intervals across 100 Monte Carlo iterations, consistently identifying top transmission pairs.
• * Sensitivity analyses confirmed that both genetic and epidemiological components meaningfully contribute to the inferred transmission structure.
• * Haplotype-derived transmission networks were significantly more informative than consensus-based networks, revealing more transmission edges for DENV2 and DENV3.

Conclusions:

• * DENcode is a robust framework for exploring dengue transmission dynamics within communities, offering a probabilistic network output.
• * The integration of pathogen genetic similarity and epidemiological parameters significantly enhances the resolution of transmission networks.
• * The framework aids in identifying key transmission events and individuals, crucial for targeted public health interventions against dengue.