Scholarlink

<p>Reconstructing individual-level antibody dynamics from sparse and irregularly sampled serological data remains challenging, particularly under hybrid immunity shaped by repeated vaccination and infection. We developed a trajectory reconstruction framework that integrates a piecewise exponential kinetic model with a nonlinear mixed-effects approach to infer event-specific antibody kinetics from longitudinal real-world data. The model represents each individual’s anti-Spike binding antibody response as a continuous sequence of boosting and waning phases across primary vaccination, second dose, booster vaccination, and breakthrough infection, while jointly estimating population-level typical kinetics and subject-specific deviations. We applied the framework to longitudinal antibody measurements from 380 healthcare workers in South Korea with heterogeneous sampling schedules, vaccine regimens, and immune histories. The framework enabled reconstruction of individual antibody trajectories from irregular observations, with a median individual-level Standardized Root Mean Square Error (SRMSE) calculated on the <InlineEquation ID="IEq1"><EquationSource Format="TEX">\(\log _{10}\)</EquationSource></InlineEquation>-scale of 0.036. The median <InlineEquation ID="IEq2"><EquationSource Format="TEX">\(\log _{10}\)</EquationSource></InlineEquation>-scale Mean Absolute Percentage Error (MAPE) was 3.51%; elevated log-scale MAPE values were observed in a subset of participants, while log-scale SRMSE remained relatively stable across individuals. These reconstructed trajectories enabled characterization of individual- and event-specific response patterns that are difficult to capture using snapshot-based or population-level approaches. Our findings support trajectory-based modeling for evaluating immune dynamics under hybrid immunity.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Individual-level reconstruction of longitudinal anti-spike binding antibody kinetics under hybrid immunity

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年8月5日

查看全文
  • Young Kim,
  • Jae-Hoon Ko,
  • Yun Bae Kim,
  • Sunmi Lee

摘要

Reconstructing individual-level antibody dynamics from sparse and irregularly sampled serological data remains challenging, particularly under hybrid immunity shaped by repeated vaccination and infection. We developed a trajectory reconstruction framework that integrates a piecewise exponential kinetic model with a nonlinear mixed-effects approach to infer event-specific antibody kinetics from longitudinal real-world data. The model represents each individual’s anti-Spike binding antibody response as a continuous sequence of boosting and waning phases across primary vaccination, second dose, booster vaccination, and breakthrough infection, while jointly estimating population-level typical kinetics and subject-specific deviations. We applied the framework to longitudinal antibody measurements from 380 healthcare workers in South Korea with heterogeneous sampling schedules, vaccine regimens, and immune histories. The framework enabled reconstruction of individual antibody trajectories from irregular observations, with a median individual-level Standardized Root Mean Square Error (SRMSE) calculated on the \(\log _{10}\)-scale of 0.036. The median \(\log _{10}\)-scale Mean Absolute Percentage Error (MAPE) was 3.51%; elevated log-scale MAPE values were observed in a subset of participants, while log-scale SRMSE remained relatively stable across individuals. These reconstructed trajectories enabled characterization of individual- and event-specific response patterns that are difficult to capture using snapshot-based or population-level approaches. Our findings support trajectory-based modeling for evaluating immune dynamics under hybrid immunity.