×

Stochastic epidemic models inference and diagnosis with Poisson random measure data augmentation. (English) Zbl 1475.92183

Summary: We present a new Bayesian inference method for compartmental models that takes into account the intrinsic stochasticity of the process. We show how to formulate a SIR-type Markov jump process as the solution of a stochastic differential equation with respect to a Poisson random measure (PRM), and how to simulate the process trajectory deterministically from a parameter value and a PRM realization.
This forms the basis of our data augmented MCMC, which consists of augmenting parameter space with the unobserved PRM value. The resulting simple Metropolis-Hastings sampler acts as an efficient simulation-based inference method, that can easily be transferred from model to model.
Compared with a recent data augmentation method based on Gibbs sampling of individual infection histories, PRM-augmented MCMC scales much better with epidemic size and is far more flexible. It is also found to be competitive with Particle MCMC for moderate epidemics when using approximate simulations.
PRM-augmented MCMC also yields a posteriori estimates of the PRM, that represent process stochasticity, and which can be used to validate the model. A pattern of deviation from the PRM prior distribution will indicate that the model underfits the data and help to understand the cause. We illustrate this by fitting a non-seasonal model to some simulated seasonal case count data.
Applied to the zika epidemic of 2013 in French Polynesia, our approach shows that a simple SEIR model cannot correctly reproduce both the initial sharp increase in the number of cases as well as the final proportion of seropositive.
PRM augmentation thus provides a coherent story for stochastic epidemic model inference, where explicitly inferring process stochasticity helps with model validation.

MSC:

92D30 Epidemiology
60H10 Stochastic ordinary differential equations (aspects of stochastic analysis)
60G57 Random measures

Software:

BEAST; pomp; ramcmc
PDFBibTeX XMLCite
Full Text: DOI arXiv

References:

[1] O’Neill, Philip D., A tutorial introduction to Bayesian inference for stochastic epidemic models using Markov chain Monte Carlo methods, Math. Biosci., 180, 1-2, 103-114 (2002) · Zbl 1021.62094
[2] Wallace, E. W.J.; Gillespie, D. T.; Sanft, K. R.; Petzold, L. R., Linear noise approximation is valid over limited times for any chemical system that is sufficiently large, IET Syst. Biol., 6, 4, 102-115 (2012)
[3] Fearnhead, Paul; Giagos, Vasilieos; Sherlock, Chris, Inference for reaction networks using the linear noise approximation, Biometrics, 70, 2, 457-466 (2014) · Zbl 1419.62346
[4] Andrieu, Christophe; Doucet, Arnaud; Holenstein, Roman, Particle Markov chain Monte Carlo methods: particle Markov chain Monte Carlo methods, J. R. Stat. Soc. Ser. B Stat. Methodol., 72, 3, 269-342 (2010) · Zbl 1411.65020
[5] Moral, Pierre Del; Jasra, Ajay; Lee, Anthony; Yau, Christopher; Zhang, Xiaole, The alive particle filter and its use in particle Markov chain Monte Carlo, Stoch. Anal. Appl., 33, 6, 943-974 (2015) · Zbl 1337.82019
[6] Cauchemez, S.; Carrat, F.; Viboud, C.; Valleron, A. J.; Boëlle, P. Y., A Bayesian MCMC approach to study transmission of influenza: Application to household longitudinal data, Stat. Med., 23, 22, 3469-3487 (2004)
[7] Drummond, Alexei J.; Rambaut, Andrew, BEAST: Bayesian evolutionary analysis by sampling trees, BMC Evol. Biol., 7, 1, 214 (2007)
[8] Jewell, Chris P.; Kypraios, Theodore; Neal, Peter; Roberts, Gareth O., Bayesian analysis for emerging infectious diseases, Bayesian Anal., 4, 3, 465-496 (2009) · Zbl 1330.62395
[9] Bouckaert, Remco; Heled, Joseph; Kühnert, Denise; Vaughan, Tim; Wu, Chieh-Hsi; Xie, Dong; Suchard, Marc A.; Rambaut, Andrew; Drummond, Alexei J., BEAST 2: A software platform for Bayesian evolutionary analysis, PLoS Comput. Biol., 10, 4, Article e1003537 pp. (2014)
[10] Ikeda, Nobuyuki; Watanabe, Shinzo, (Stochastic Differential Equations and Diffusion Processes. Stochastic Differential Equations and Diffusion Processes, North-Holland Mathematical Library, no. 24 (1989), North-Holland [u.a.]: North-Holland [u.a.] Amsterdam) · Zbl 0684.60040
[11] Neal, Peter; Roberts, Gareth, A case study in non-centering for data augmentation: Stochastic epidemics, Stat. Comput., 15, 4, 315-327 (2005)
[12] Xiang, Fei; Neal, Peter, Efficient MCMC for temporal epidemics via parameter reduction, Comput. Statist. Data Anal., 80, 240-250 (2014) · Zbl 1506.62194
[13] Neal, Peter; Huang, Chien Lin Terry, Forward simulation Markov chain Monte Carlo with applications to stochastic epidemic models, Scand. J. Stat., 42, 2, 378-396 (2015) · Zbl 1376.62086
[14] Fintzi, Jonathan; Cui, Xiang; Wakefield, Jon; Minin, Vladimir N., Efficient data augmentation for fitting stochastic epidemic models to prevalence data, J. Comput. Graph. Statist., 26, 4, 918-929 (2017)
[15] Rao, Vinayak; Teh, Yee Whye, Fast MCMC sampling for Markov jump processes and extensions, J. Mach. Learn. Res., 14, 26 (2013) · Zbl 1318.60078
[16] Golightly, Andrew; Wilkinson, Darren J., Bayesian inference for Markov jump processes with informative observations, Stat. Appl. Genet. Mol. Biol., 14, 2, 169-188 (2015) · Zbl 1311.92089
[17] Zhang, Boqian; Rao, Vinayak, Efficient parameter sampling for Markov jump processes (2017), arXiv:1704.02369 [stat], April
[18] Metropolis, Nicholas; Rosenbluth, Arianna W.; Rosenbluth, Marshall N.; Teller, Augusta H.; Teller, Edward, Equation of state calculations by fast computing machines, J. Chem. Phys., 21, 6, 1087-1092 (1953) · Zbl 1431.65006
[19] Hastings, W. K., Monte Carlo sampling methods using Markov chains and their applications, Biometrika, 57, 1, 97-109 (1970) · Zbl 0219.65008
[20] Jensen, Arne, Markoff chains as an aid in the study of Markoff processes, Scand. Actuar. J., 1953, Suppl. 1, 87-91 (1953) · Zbl 0051.35607
[21] King, Aaron A.; Nguyen, Dao; Ionides, Edward L., Statistical inference for partially observed Markov processes via the R package pomp, J. Stat. Softw., 69, 1, 1-43 (2016)
[22] Lau, M. S.Y.; Marion, G.; Streftaris, G.; Gibson, G. J., New model diagnostics for spatio-temporal systems in epidemiology and ecology, J. R. Soc. Interface, 11, 93 (2014), 20131093
[23] Ethier, Stewart N.; Kurtz, Thomas G., Markov Processes: Characterization and Convergence (2009), John Wiley & Sons · Zbl 0592.60049
[24] Dureau, Joseph; Ballesteros, Sébastien; Bogich, Tiffany, SSM: inference for time series analysis with state space models (2013), arXiv:1307.5626 [stat], July
[25] Bansaye, Vincent; Méléard, Sylvie, Stochastic Models for Structured Populations (2015), Springer · Zbl 1333.92004
[26] Papaspiliopoulos, Omiros; Roberts, Gareth O.; Sköld, Martin, A general framework for the parametrization of hierarchical models, Statist. Sci., 22, 1, 59-73 (2007) · Zbl 1246.62195
[27] Xavier Leroy, OCaml, an industrial strength programming language.
[28] Vihola, Matti, Robust adaptive Metropolis algorithm with coerced acceptance rate, Stat. Comput., 22, 5, 997-1008 (2012) · Zbl 1252.65024
[29] Vats, Dootika; Flegal, James M.; Jones, Galin L., Multivariate output analysis for Markov chain Monte Carlo, Biometrika, 106, 2, 321-337 (2019) · Zbl 1434.62100
[30] Kucharski, Adam J.; Funk, Sebastian; Eggo, Rosalind M.; Mallet, Henri-Pierre; Edmunds, W. John; Nilles, Eric J., Transmission dynamics of zika virus in island populations: a modelling analysis of the 2013-14 French polynesia outbreak, PLoS Neglect. Trop. Dis., 10, 5 (2016), e0004726
[31] Cauchemez, Simon; Besnard, Marianne; Bompard, Priscillia; Dub, Timothée; Guillemette-Artur, Prisca; Eyrolle-Guignot, Dominique; Salje, Henrik; Kerkhove, Maria D. Van; Abadie, Véronique; Garel, Catherine; Fontanet, Arnaud; Mallet, Henri-Pierre, Association between Zika virus and microcephaly in French Polynesia, 2013-15: A retrospective study, Lancet, 387, 10033, 2125-2132 (2016)
[32] Champagne, Clara; Salthouse, David Georges; Paul, Richard; Cao-Lormeau, Van-Mai; Roche, Benjamin; Cazelles, Bernard, Structure in the variability of the basic reproductive number (R0) for Zika epidemics in the Pacific Islands, eLife, 5, Article e19874 pp. (2016)
[33] Mallet, H. P.; Vial, Anne-Laure; Musso, Didier, Bilan de l’epidemie a Virus Zika En Polynesie Francaise, 2013-2014, Bull. Inf. Sanitaires Épidémiol. Stat., 20-21 (2015)
[34] Aubry, Maite; Teissier, Anita; Huart, Michael; Merceron, Sébastien; Vanhomwegen, Jessica; Roche, Claudine; Vial, Anne-Laure; Teururai, Sylvianne; Sicard, Sébastien; Paulous, Sylvie; Desprès, Philippe; Manuguerra, Jean-Claude; Mallet, Henri-Pierre; Musso, Didier; Deparis, Xavier; Cao-Lormeau, Van-Mai, Zika virus seroprevalence, French polynesia, 2014-2015, Emerg. Infect. Diseases, 23, 4, 669-672 (2017)
[35] Cazelles, Bernard; Champagne, Clara; Dureau, Joseph, Accounting for non-stationarity in epidemiology by embedding time-varying parameters in stochastic models, PLoS Comput. Biol., 14, 8 (2018), e1006211
[36] Haario, Heikki; Saksman, Eero; Tamminen, Johanna, An adaptive Metropolis algorithm, Bernoulli, 7, 2, 223-242 (2001) · Zbl 0989.65004
[37] Roberts, Gareth O.; Rosenthal, Jeffrey S., Examples of adaptive MCMC, J. Comput. Graph. Statist., 18, 2, 349-367 (2009)
[38] Pooley, C. M.; Bishop, S. C.; Marion, G., Using model-based proposals for fast parameter inference on discrete state space, continuous-time Markov processes, J. R. Soc. Interface, 12, 107 (2015), 20150225
[39] Fintzi, Jonathan; Wakefield, Jon; Minin, Vladimir N., A linear noise approximation for stochastic epidemic models fit to partially observed incidence counts (2020), arXiv:2001.05099 [q-bio, stat], January
[40] Lau, Max S. Y.; Grenfell, Bryan T.; Worby, Colin J.; Gibson, Gavin J., Model diagnostics and refinement for phylodynamic models, PLoS Comput. Biol., 15, 4, Article e1006955 pp. (2019)
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.