bdct

bdct integrates contact tracing into Birth-Death with Contact Tracing (BD-CT) phylodynamic models to infer epidemiological parameters from pathogen phylogenetic trees.

Key Features:

• Extension of MTBD models: Extends multi-type birth-death (MTBD) models by integrating contact tracing mechanisms (MTBD-CT) to model transmission dynamics under contact-tracing interventions.
• Simulation capabilities: Includes a simulator that generates phylogenetic trees under both MTBD and MTBD-CT frameworks for model testing and validation.
• Non-parametric contact-tracing detection test: Implements a novel non-parametric test to detect the presence of contact tracing in pathogen phylogenetic trees with demonstrated high specificity and sensitivity on simulated datasets.
• Closed-form likelihood for BD-CT(1): Provides a closed-form solution for the likelihood function under the BD-CT(1) model (only the last contact notified), enabling maximum-likelihood estimation of model parameters.
• Parameter inference and confidence intervals: Enables maximum-likelihood estimation of epidemiological parameters and their confidence intervals directly from phylogenetic trees using the BD-CT(1) likelihood.
• Bias reduction in parameter estimation: Accounts for contact tracing to prevent biases such as overestimation of the becoming-non-infectious rate that occur when contact tracing is ignored.
• Application to real datasets: Applied to detect contact tracing in HIV-1 B epidemics in Zurich and the UK.

Scientific Applications:

• Detection of contact tracing in phylogenies: Detects signatures of contact tracing within pathogen phylogenetic trees, as demonstrated for HIV-1 B in Zurich and the UK.
• Improved epidemiological parameter estimation: Produces more accurate estimates of parameters such as the becoming-non-infectious rate by accounting for contact tracing.
• Support for epidemiological studies and public-health assessment: Informs phylodynamic analyses and evaluation of contact-tracing effects in diseases where contact tracing is a key intervention.

Methodology:

Extends MTBD to MTBD-CT models; simulates phylogenetic trees under MTBD and MTBD-CT; implements a non-parametric test for contact-tracing detection; derives a closed-form likelihood for BD-CT(1) and performs maximum-likelihood parameter estimation with confidence intervals evaluated on simulated datasets.