Estimating time since infection in early homogeneous HIV-1 samples using a poisson model

Area of Science:

• Virology
• Evolutionary Biology
• Computational Biology

Background:

• Genetic bottlenecks are common in human immunodeficiency virus (HIV) transmission, leading to homogeneous infections initiated by a single strain.
• Viral populations grow exponentially early in infection, before the host immune response significantly impacts viral evolution.
• Existing methods for estimating evolutionary parameters, such as Bayesian phylogenetic inference (e.g., BEAST), rely on simulating complex genealogies.

Purpose of the Study:

• To develop and present a novel web tool for analyzing genetic diversity in acutely HIV-1 infected patients.
• To provide a computationally efficient alternative to existing methods for estimating evolutionary and demographic parameters in early HIV infections.
• To identify deviations from a neutral growth model as indicators of heterogeneous infections or the onset of selection.

Main Methods:

• The tool compares genetic diversity measures in patient samples to a model of neutral growth, assuming random mutation accumulation in homogeneous infections.
• It models early HIV-1 infection as a scenario of homogeneous initiation followed by exponential viral growth prior to host immune selection.
• Statistical tests are performed on the Hamming distance frequency distribution, yielding summary statistics like the mean of the best-fitting Poisson distribution and goodness-of-fit p-values.

Main Results:

• The developed web tool successfully analyzes genetic diversity in acutely HIV-1 infected individuals.
• The neutral growth model accurately describes approximately 80% of early sexual HIV-1 transmissions when samples are collected sufficiently early.
• Deviations from the model suggest either heterogeneous infection origins or the emergence of immune-driven selection.

Conclusions:

• The tool employs a forward-time approach, computationally more efficient than backward-time coalescent methods, for estimating the time to the most recent common ancestor under specific conditions.
• It provides a rapid analysis, completing within minutes and capable of handling large datasets from ultradeep pyrosequencing.
• The web tool is accessible on the Los Alamos National Laboratory (LANL) website, facilitating its use in HIV research.