Sample efficient likelihood-free inference for virus dynamics with different types of experiments

2025-08-14

AI Summary

This study addresses the challenge of parameter inference in viral dynamics models when an explicit likelihood function is unavailable. Methodologically, we propose an enhanced Bayesian Optimization for Likelihood-Free Inference (BOLFI) framework: (1) replacing the conventional Gaussian process regression surrogate with a Gaussian process classifier to markedly improve robustness against outliers and noise in experimental data; and (2) introducing a differentiable, cross-modal discrepancy metric that enables, for the first time, joint inference from heterogeneous virological data modalities—including viral titers, cellular infection rates, and fluorescence imaging. Evaluated on real influenza virus data, our method accelerates inference by 3–5× over standard MCMC while yielding biologically plausible parameter estimates and well-calibrated posterior uncertainty quantification. The core contribution lies in eliminating reliance on hand-crafted likelihoods, unifying inference across diverse dynamical data types, and establishing an efficient, interpretable paradigm for modeling complex pathogens.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Search and Optimization: Sampling/Simulation-based Search
• Reasoning under Uncertainty: Stochastic Optimization

Application Category

• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs

Abstract

This study applied Bayesian optimization likelihood-free inference(BOLFI) to virus dynamics experimental data and efficiently inferred the model parameters with uncertainty measure. The computational benefit is remarkable compared to existing methodology on the same problem. No likelihood knowledge is needed in the inference. Improvement of the BOLFI algorithm with Gaussian process based classifier for treatment of extreme values are provided. Discrepancy design for combining different forms of data from completely different experiment processes are suggested and tested with synthetic data, then applied to real data. Reasonable parameter values are estimated for influenza A virus data.

Problem

• Efficiently infer virus dynamics model parameters without likelihood knowledge
• Combine different experiment data types for accurate parameter estimation
• Improve Bayesian optimization for handling extreme values in virus data

Innovation

• Bayesian optimization for likelihood-free inference
• Gaussian process classifier for extreme values
• Discrepancy design for diverse experiment data