Markov chain Monte Carlo without likelihoods
Open accessThe DOI matches its Crossref record; 20 of the 20 references it cites were found in a registry or an index.
Abstract
Many stochastic simulation approaches for generating observations from a posterior distribution depend on knowing a likelihood function. However, for many complex probability models, such likelihoods are either impossible or computationally prohibitive to obtain. Here we present a Markov chain Monte Carlo method for generating observations from a posterior distribution without the use of likelihoods. It can also be used in frequentist applications, in particular for maximum-likelihood estimation. The approach is illustrated by an example of ancestral inference in population genetics. A number of open problems are highlighted in the discussion.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“The greatest computational advantage of the original ABC rejection algorithm over Monte Carlo based ABC methods is that it generates independent samples and therefore readily lends itself to ‘embarrassingly’ parallel computation [86]. The computation time thus scales linearly to the number of processors available.”
From Integrating Crop Growth Models with Whole Genome Prediction through Approximate Bayesian Computation (2015) · cites this paper as [86]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“If ϵ is sufficiently small then this distribution will be a good approximation for the ‘true’ posterior distribution, P(θ|D0).”
From Simulation-based model selection for dynamical systems in systems and population biology (2010) · cites this paper as Marjoram et al., 2003
Experimental — accuracy not yet measuredThe check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“Moreover, related ideas are already applied in the context of generalized linear mixed models [16], unobserved data points [17].”
From Likelihood based observability analysis and confidence intervals for predictions of dynamic models (2012) · cites this paper as [17]
Experimental — accuracy not yet measuredIn the citing paper
“The price to pay is the generation of dependent outcomes, and the risk of getting stuck in regions of low probability of the state space for long periods of time.”
From Reverse engineering and identification in systems biology: strategies, perspectives and challenges (2014) · cites this paper as [193]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Equation of State Calculations by Fast Computing Machines
Metropolis, N., Rosenbluth, A. W., Rosenbluth, M. N., Teller, A. H., & Teller, E. (1953) · The Journal of Chemical Physics · doi:10.1063/1.1699114 · Cited by 37,192
Gelman, A., Carlin, J. B., Stern, H. S., & Rubin, D. B. (2003) · doi:10.1201/9780429258480 · Cited by 27,276
Monte Carlo sampling methods using Markov chains and their applications.
Hastings, W. K. (1970) · Biometrika · doi:10.1093/biomet/57.1.97 · Cited by 15,195
Markov Chain Monte Carlo in Practice
(1995) · Chapman & Hall/CRC eBooks · doi:10.1201/b14835 · Cited by 7,757
Approximate Bayesian computation in population genetics
Beaumont, M. A., Zhang, W., & Balding, D. J. (2002) · Genetics · doi:10.1093/genetics/162.4.2025 · Cited by 2,932
Ripley, B. D. (1987) · Wiley Series in Probability and Statistics · doi:10.1002/9780470316726 · Cited by 2,241
Bayes and Empirical Bayes Methods for Data Analysis
(2001) · Technometrics · doi:10.1198/tech.2001.s608 · Cited by 2,098
On the genealogy of large populations
Kingmán, J. F. C. (1982) · Journal of Applied Probability · doi:10.1017/s0021900200034446 · Cited by 1,628
Population growth of human Y chromosomes: a study of Y chromosome microsatellites
Pritchard, J. K., Seielstad, M. T., Perez-Lezaun, A., & Feldman, M. W. (1999) · Molecular Biology and Evolution · doi:10.1093/oxfordjournals.molbev.a026091 · Cited by 1,080
Bayes and Empirical Bayes Methods for Data Analysis, Second Edition
Carlin, B., & Louis, T. (2000) · Chapman & Hall/CRC Texts in Statistical Science · doi:10.1201/9781420057669 · Cited by 959
Research works printed but not linked: 4
- Kingman, J. F. C. (1982) J. Appl. Prob. 19A, 27–43. [Google Scholar] · doi:10.2307/3213548 · Paper page
- Stephens, M. (2001) in Handbook of Statistical Genetics, eds. Balding, D. J., Bishop, M. J. & Cannings, C. (Wiley, New York), pp. 213–238.
- Plagnol, V. & Tavaré, S. (2004) in Proceedings of the 5th International Conference on Monte Carlo and Quasi-Monte Carlo Methods in Scientific Computing, ed. Niederreiter, H. (Springer, Heidelberg).
- Cabrera, J. & Yohai, V. J. (1999) A New Computational Approach for Bayesian and Robust Bayesian Statistical Analysis, www.rci.rutgers.edu/~cabrera/pap/vic.pdf, preprint.
Cited by
Items 301–400 of 1,000
Mardulyn, P., Goffredo, M., Conte, A., Hendrickx, G., Meiswinkel, R., Balenghien, T., Sghaier, S., Lohr, Y., & Gilbert, M. (2013) · Molecular Ecology · doi:10.1111/mec.12264 · Cited by 41
Approximate Bayesian forecasting
Frazier, D. T., Maneesoonthorn, W. O., Martin, G. M., & McCabe, B. P. M. (2019) · International Journal of Forecasting · doi:10.1016/j.ijforecast.2018.08.003 · Cited by 45
“…of Algorithm 1 based on either post-sampling corrections (Beaumont et al., 2002; Blum, 2010), or the insertion of MCMC or sequential Monte Carlo steps (Marjoram et al., 2003; Sisson et al., 2007; Beaumont et al., 2009) may well improve the accuracy with which the exact posteriors are approximated.”
BackgroundFrom Summary Statistics to Gene Trees: Methods for Inferring Positive Selection
Hejase, H. A., Dukler, N., & Siepel, A. C. (2020) · Trends in Genetics · doi:10.1016/j.tig.2019.12.008 · Cited by 39
“In particular, ABC works by simulating data sets according to a prior distribution of model parameters, or in more advanced applications, from a proposal distribution in a Markov chain Monte Carlo (MCMC) setting [83,84].”
MethodWang, Y.-C., Xie, Y.-B., Zhang, T.-J., Huang, H.-C., Zhang, T., & Liu, K. (2021) · The Astrophysical Journal Supplement Series · doi:10.3847/1538-4365/abf8aa · Cited by 43
Adaptivity for ABC algorithms: the ABC-PMC scheme
Robert, C. P., Beaumont, M., Marin, J., & Cornuet, J. (2008) · Cited by 36
Chain ladder method: Bayesian bootstrap versus classical bootstrap
Peters, G. W., Wüthrich, M. V., & Shevchenko, P. V. (2010) · Insurance Mathematics and Economics · doi:10.1016/j.insmatheco.2010.03.007 · Cited by 37
“or example, a sequential Monte Carlo (SMC) based algorithms which can improve simulation efficiency can be found in Del Moral et al. [2], Sisson et al. [25], Peters et al. [17],[18] and Marjoram et al. [15]. 15 5.1. ABC methodology In this section we provide a brief description of ABC methodology, which describes a suit…”
MethodInfluentialKhazeiynasab, S. R., & Qi, J. (2021) · IEEE Transactions on Smart Grid · doi:10.1109/tsg.2021.3077734 · Cited by 36
Some models are useful, but how do we know which ones? Towards a unified Bayesian model taxonomy
Bürkner, P., Scholz, M. M., & Radev, S. T. (2023) · Statistics Surveys · doi:10.1214/23-ss145 · Cited by 51
“More sophisticated ABC algorithms, such as ABC-SMC [276, 166] or ABC-MCMC [194, 290] alleviate some of these issues and inherit the convergence diagnostics of SMC and MCMC.”
MethodApproximate Bayesian computing for spatial extremes
Erhardt, R. J., & Smith, R. L. (2011) · Computational Statistics & Data Analysis · doi:10.1016/j.csda.2011.12.003 · Cited by 45
Likelihood Methods for CMB Experiments
Gerbino, M., Lattanzi, M., Migliaccio, M., Pagano, L., Salvati, L., Colombo, L. P. L., GRUPPUSO, A., Natoli, P., & Polenta, G. (2020) · Frontiers in Physics · doi:10.3389/fphy.2020.00015 · Cited by 42
“ABC methods have been initially employed in various fields as a way to bypass the evaluation of the likelihood function with the use of simulated data [72, 73, 74, 75, 76, 77, 78, 79, 80, 81].”
MethodThe use of a single pseudo-sample in approximate Bayesian computation
Bornn, L., Pillai, N. S., Smith, A. D., & Woodard, D. B. (2016) · Statistics and Computing · doi:10.1007/s11222-016-9640-7 · Cited by 47
“For any unbiased, nonnegative estimator μ̂(θ) of πK(θ|yobs) and any reversible transition kernel q on Θ, Algorithm 2 gives a Markov chain Monte Carlo (MCMC) version that is based on the pseudo-marginal algorithm of (Andrieu and Roberts, 2009) (see also (Marjoram et al., 2003; Wilkinson, 2013; Andrieu and Vihola, 2014)…”
MethodBayesian Phylogenetic Inference Using a Combinatorial Sequential Monte Carlo Method
Wang, L., Bouchard‐Côté, A., & Doucet, A. (2015) · Journal of the American Statistical Association · doi:10.1080/01621459.2015.1054487 · Cited by 47
“This is because a richer class of trees is needed in phylogenetics models, whereas related SMC work on population genetics typically focuses on coalescent models (Griffiths and Tavaré 1996; Liu 2001; Beaumont et al. 2002; Marjoram et al. 2002; Iorio and Griffiths 2004; Paul et al. 2011).”
BackgroundBayesian Learning from Marginal Data in Bionetwork Models
Bonassi, F. V., You, L., & West, M. (2011) · Statistical Applications in Genetics and Molecular Biology · doi:10.2202/1544-6115.1684 · Cited by 48
“Direct rejection-based ABC methods, ABC-MCMC or sequential Monte Carlo variants (Marjoram et al., 2003, Sisson et al., 2007, Blum and François, 2010, Nunes and Balding, 2010) are useful in such contexts where there is an ability to simulate a model but not evaluate likelihood functions.”
An approximate likelihood perspective on ABC methods
Karabatsos, G., & Leisen, F. (2018) · Statistics Surveys · doi:10.1214/18-ss120 · Cited by 43
“This is also true for kernel ABC (Section 4.2) and R-ABC-MCMC (Section 5) [38].”
BackgroundInfluentialVariational Implicit ProcessesPreprint
Ma, C., Li, Y., & Hernández-Lobato, J. M. (2018) · arXiv (Cornell University) · doi:10.17863/cam.42186 · Cited by 81
Fisher, P., Spencer, H., THOMAS‐OATES, J. E., Wood, A. J., & Ungár, D. (2019) · Cell Reports · doi:10.1016/j.celrep.2019.03.107 · Cited by 38
“To determine a set of appropriate rates for the SSA based on the relative effective enzymatic rates of the glycan-modifying enzymes and their unique localizations within the Golgi stack, our computational method is parameterized using a Bayesian framework (Figure S1) (Marjoram et al., 2003).”
MethodChallenger, J. D., Gonçalves, B. P., Bradley, J., Bruxvoort, K. J., Tiono, A. B., DRAKELEY, C. J., Bousema, T., Ghani, A. C., & Okell, L. (2019) · BMJ Global Health · doi:10.1136/bmjgh-2019-001856 · Cited by 36
Avecilla, G., Chuong, J. N., Li, F., Sherlock, G., Gresham, D., & Ram, Y. (2022) · PLoS Biology · doi:10.1371/journal.pbio.3001633 · Cited by 46
Approximating Bayesian Inference through Model Simulation
Turner, B. M., & Zandt, T. V. (2018) · Trends in Cognitive Sciences · doi:10.1016/j.tics.2018.06.003 · Cited by 42
Control of equine influenza: scenario testing using a realistic metapopulation model of spread
Baguelin, M., Newton, J. R., Demiris, N., Daly, J. M., Mumford, J. A., & Wood, J. L. N. (2009) · Journal of The Royal Society Interface · doi:10.1098/rsif.2009.0030 · Cited by 41
“It is then possible to estimate lL and lG by considering that the frequency of occurrences respecting both conditions is a simulated likelihood of the given pair of parameters, in a similar way as for approximate Bayesian computation (Beaumont et al. 2002; Marjoram et al. 2003).”
MethodMarion, G., McInerny, G., Pagel, J., Catterall, S., Cook, A. R., Hartig, F., & O’Hara, R. B. (2012) · Journal of Biogeography · doi:10.1111/j.1365-2699.2012.02772.x · Cited by 37
“require a switch from the explicit likelihood approach adopted in this paper to likelihood-free approaches (Marjoram et al., 2003), approximate Bayesian computation (see”
Seeing Is Believing: Quantifying Is Convincing: Computational Image Analysis in Biology
Sbalzarini, I. F. (2016) · Advances in anatomy, embryology and cell biology · doi:10.1007/978-3-319-28549-8_1 · Cited by 37
Marchand, P., Boenke, M. C., & Green, D. M. (2017) · Ecological Modelling · doi:10.1016/j.ecolmodel.2017.06.025 · Cited by 35
“Alternative ABC algorithms use Markov chain Monte Carlo or sequential Monte Carlo (a.k.a. particle filter) methods to gradually concentrate the sampling effort in the areas of high-agreement between simulated and observed statistics (Marjoram et al., 2003; Sisson et al., 2007).”
MethodA comparison of likelihood-free methods with and without summary statistics
Drovandi, C., & Frazier, D. T. (2022) · Statistics and Computing · doi:10.1007/s11222-022-10092-4 · Cited by 50
“Problems with performing model choice on the basis of summary statistics have been well documented (Robert et al., 2011; Marin et al., 2013).”
BackgroundEfficient likelihood-free Bayesian Computation for household epidemics
Neal, P. J. (2010) · Statistics and Computing · doi:10.1007/s11222-010-9216-x · Cited by 40
“For MCMC-ABC, see (Marjoram et al. 2003) for details, the average acceptance probability, P MCMC , of a proposed simulation is where q(φ | θ) is the probability density function for proposing the move from θ to φ and α(θ,φ) is the acceptance probability of the proposed move given that where π q (φ…”
BackgroundMethodRasmussen, R., & Hamilton, G. S. (2011) · Environmental Modelling & Software · doi:10.1016/j.envsoft.2011.10.005 · Cited by 40
“In Algorithm 2 Rejection ABC 1: Draw θ so that θ ∼ π(·) 2: Simulate Ds so that Ds ∼ φ(θ) 3: Accept θ if d(S(Ds), S(Do)) ≤ ǫ; otherwise go to 1 these cases, importance sampling may be applied in a likelihood-free Markov Chain Monte Carlo (MCMC) search (see Marjoram et al., 2003).”
BackgroundMethodInfluentialPhylodynamic Inference with Kernel ABC and Its Application to HIV Epidemiology
Poon, A. F. Y. (2015) · Molecular Biology and Evolution · doi:10.1093/molbev/msv123 · Cited by 33
“The kernel function was employed to sample parameters from the posterior distribution in the framework of MCMC for ABC (ABC–MCMC; Marjoram et al. 2003).”
MethodInfluentialMacLean, A. L., Celso, C. L., & Stumpf, M. P. H. (2013) · Journal of The Royal Society Interface · doi:10.1098/rsif.2012.0968 · Cited by 37
Importance sampling type estimators based on approximate marginal Markov chain Monte Carlo
Vihola, M., Helske, J., & Franks, J. (2020) · Scandinavian Journal of Statistics · doi:10.1111/sjos.12492 · Cited by 34
Bayesian model comparison with un-normalised likelihoods
Everitt, R. G., Johansen, A. M., Rowing, E., & Evdemon-Hogan, M. (2016) · Statistics and Computing · doi:10.1007/s11222-016-9629-2 · Cited by 39
“Since a rejection sampler is not available in many cases, we do not pursue this approach further.”
Bayesian Active Learning for Posterior Estimation - IJCAI-15 Distinguished Paper.
Kandasamy, K., Schneider, J., & Póczos, B. (2015) · International Joint Conference on Artificial Intelligence · Cited by 32
Eriksson, O., Jauhiainen, A., Sasane, S. M., Kramer, A., Nair, A. G., Sartorius, C., & Kotaleski, J. H. (2018) · Bioinformatics · doi:10.1093/bioinformatics/bty607 · Cited by 37
“A Markov Chain Monte Carlo (MCMC) approach is used for the ABC sampling (ABC-MCMC) on each dataset in the sequence, and after the final step of the sequence, we also boost the posterior sample by a simple rejection sampling from the final copula.”
MethodInfluentialEstimating DSGE Models: Recent Advances and Future Challenges
Fernández‐Villaverde, J., & Guerron-Quintana, P. A. (2021) · Annual Review of Economics · doi:10.1146/annurev-economics-081020-044812 · Cited by 47
Järvenpää, M., Gutmann, M. U., Vehtari, A., & Marttinen, P. (2018) · The Annals of Applied Statistics · doi:10.1214/18-aoas1150 · Cited by 37
“To speed up ABC inference, several sampling-based algorithms have been proposed [Marjoram et al. (2003), Sisson, Fan and Tanaka (2007), Beaumont et al. (2009), Toni et al. (2009), Drovandi and Pettitt (2011), Del Moral, Doucet and Jasra (2012), Lenormand, Jabot and Deffuant (2013)].”
MethodA structural model of segregation in social networks
Mele, A. (2010) · doi:10.1920/wp.cem.2010.3210 · Cited by 32
Practical bounds on the error of Bayesian posterior approximations: A nonasymptotic approachPreprint
Huggins, J. H., Campbell, T., Kasprzak, M. J., & Broderick, T. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1809.09505 · Cited by 32
Jiang, Y., Guo, J., Zhao, D., & Li, Y. (2024) · Energy · doi:10.1016/j.energy.2024.132774 · Cited by 31
Approximate Bayesian Computation for a Class of Time Series Models
Jasra, A. (2015) · International Statistical Review · doi:10.1111/insr.12089 · Cited by 32
“he space Yn: ˇ(;u 1:njy 1:n) / Yn i=1 I fu i2Y:d(u i;y i)<g(u i)p (u i) ˇ(): In Algorithm 2 we present a natural Metropolis-Hastings (M-H) proposal (which is essentially the MCMC method of [56]) that could be used to sample from ˇ(;u 1:njy 1:n). The one-step transition kernel of the MCMC chain is usually described…”
MethodHe, J., Guan, X., & Jha, R. (2016) · IEEE Transactions on Reliability · doi:10.1109/tr.2016.2604121 · Cited by 31
“, particle filtering) [31]–[33].”
Query efficient posterior estimation in scientific experiments via Bayesian active learning
Kandasamy, K., Schneider, J., & Póczos, B. (2016) · Artificial Intelligence · doi:10.1016/j.artint.2016.11.002 · Cited by 31
“There are several variants of ABC [18, 22].”
MethodSelecting Summary Statistics in Approximate Bayesian Computation for Calibrating Stochastic Models
Burr, T. L., & Skurikhin, A. N. (2013) · BioMed Research International · doi:10.1155/2013/210646 · Cited by 35
Patterns of Neutral Genetic Variation on Recombining Sex Chromosomes
Kirkpatrick, M., Guerrero, R. F., & Scarpino, S. V. (2010) · Genetics · doi:10.1534/genetics.109.113555 · Cited by 37
Martino, L., & Elvira, V. (2017) · Wiley eBooks · doi:10.1002/9781118445112.stat07951 · Cited by 39
Statistical Implementations of Agent‐Based Demographic Models
Hooten, M. B., Wikle, C. K., & Schwob, M. R. (2020) · International Statistical Review · doi:10.1111/insr.12399 · Cited by 34
“The ABC approach introduces robustness in the likelihood approximation by using a suitably chosen discrepancy measure instead of a likelihood in an MCMC algorithm (e.g. Beaumont et al., 2002; Marjoram et al., 2003).”
MethodReichert, P., Ammann, L., & Fenicia, F. (2020) · Water Resources Research · doi:10.1029/2020wr028400 · Cited by 33
“The simplest algorithm just consists of sampling from the prior and accepting the sample if the model prediction is closer to the data than some pre-defined tolerance (Marjoram et al., 2003).”
BackgroundMethodMelanoma Cell Colony Expansion Parameters Revealed by Approximate Bayesian Computation
Vo, B. N., Drovandi, C., Pettitt, A. N., & Pettet, G. J. (2015) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1004635 · Cited by 30
Model Selection in Historical Research Using Approximate Bayesian Computation
Rubio‐Campillo, X. (2016) · PLoS ONE · doi:10.1371/journal.pone.0146491 · Cited by 34
“It is not the most efficient ABC method (see [67, 68] for alternatives), but its Fig 1.”
MethodModeling and Inference for Infectious Disease Dynamics: A Likelihood-Based Approach
Bretó, C. (2018) · Statistical Science · doi:10.1214/17-sts636 · Cited by 30
“Also, Bayesian options include approximate Bayesian computation (ABC; Marjoram et al., 2003) and Particle Markov chain Monte Carlo (PMCMC; Andrieu, Doucet and Holenstein, 2010).”
Probabilistic programming: A review for environmental modellers
Krapu, C., & Borsuk, M. E. (2019) · Environmental Modelling & Software · doi:10.1016/j.envsoft.2019.01.014 · Cited by 31
Kursawe, J., Baker, R. E., & Fletcher, A. G. (2018) · Journal of Theoretical Biology · doi:10.1016/j.jtbi.2018.01.020 · Cited by 30
Ross, R. J. H., Baker, R. E., Parker, A. J., Ford, M. J., Mort, R. L., & Yates, C. A. (2017) · npj Systems Biology and Applications · doi:10.1038/s41540-017-0010-7 · Cited by 36
“Approximate Bayesian computation Here, we introduce our ABC algorithm.(19) We define M as a stochastic model that takes parameters Θ and produces data D.”
BackgroundMethodInfluentialGaussian process emulators for spatial individual‐level models of infectious disease
Pokharel, G., & Deardon, R. (2016) · Canadian Journal of Statistics · doi:10.1002/cjs.11304 · Cited by 29
“In such methods the likelihood is approximated by comparing simulated data to real data, usually via some summary metrics, within the framework of, say, an MCMC (e.g., Marjoram et al., 2003) or sequential Monte Carlo (e.g., Beaumont et al., 2009) algorithm.”
MethodComponentwise approximate Bayesian computation via Gibbs-like steps
Clarté, G., Robert, C. P., Ryder, R., & Stoehr, J. (2020) · Biometrika · doi:10.1093/biomet/asaa090 · Cited by 39
Dong, F., Hung, C., & Yang, X. (2019) · Molecular Phylogenetics and Evolution · doi:10.1016/j.ympev.2019.106671 · Cited by 30
Bakkeren, E., Herter, J. A., Huisman, J. S., Steiger, Y., Gül, E., Newson, J. P. M., Brachmann, A. O., Piel, J., Regoes, R. R., Bonhoeffer, S., et al. (2021) · eLife · doi:10.7554/elife.69744 · Cited by 29
Approximate Bayesian computation with composite score functions
Ruli, E., Sartori, N., & Ventura, L. (2015) · Statistics and Computing · doi:10.1007/s11222-015-9551-z · Cited by 35
“Nevertheless, this issue can be effectively addressed by using more advanced Monte Carlo algorithms, such as MCMCmethods (Marjoram et al. 2003), importance sampling (Fearnhead and Prangle 2012), sequential or population Monte Carlo approaches (Beaumont et al.”
MethodABC of SV: Limited information likelihood inference in stochastic volatility jump-diffusion models
Creel, M., & Kristensen, D. (2015) · Journal of Empirical Finance · doi:10.1016/j.jempfin.2015.01.002 · Cited by 32
“…problem, methods of computing estimators using likelihood-free Markov chain Monte Carlo and sequential Monte Carlo have been studied in some detail in the ABC literature (among others, see Marjoram et al., 2003; Sisson, Fan and Tanaka, 2007; Beaumont et al. 2009; Del Moral, Doucet and Jasra, 2012).”
MethodMultilevel rejection sampling for approximate Bayesian computation
Warne, D. J., Baker, R. E., & Simpson, M. J. (2018) · Computational Statistics & Data Analysis · doi:10.1016/j.csda.2018.02.009 · Cited by 39
“(2015) by replacing the Metropolis–Hasting acceptance probability in a similar way to the MCMC-ABC method (Marjoram et al., 2003).”
BackgroundMethodInfluentialAccelerating Bayesian Synthetic Likelihood With the Graphical Lasso
An, Z., South, L. F., Nott, D. J., & Drovandi, C. (2018) · Journal of Computational and Graphical Statistics · doi:10.1080/10618600.2018.1537928 · Cited by 28
“Both MCMC ABC (Marjoram et al. 2003) and MCMC BSL require the discovery of a suitable starting value as they suffer from very slow convergence when starting from a poor initial value (Lee and Łatuszy´nski 2014; Price et al. 2018).”
BackgroundMethodInfluentialGrazian, C., & Fan, Y. (2019) · Wiley Interdisciplinary Reviews Computational Statistics · doi:10.1002/wics.1486 · Cited by 33
“For example, one can define a large number of summary statistics and choose the summary statistics in Algorithm 1 as combinations of them (usually determined via regression); these approaches may be found in Nunes and Balding (2010), Fearnhead and Prangle (2012), Aeschbacher, Beaumont, and Futschik (2012), and Blum, N…”
MethodEndogenous Social Interactions with Unobserved Networks
Battaglini, M., Patacchini, E., & Rainone, E. (2021) · The Review of Economic Studies · doi:10.1093/restud/rdab058 · Cited by 28
Understanding the rheology of kaolinite clay suspensions using Bayesian inference
Ran, R., Pradeep, S., Acharige, S. K., Blackwell, B., Kammer, C., Jerolmack, D. J., & Arratia, P. E. (2022) · Journal of Rheology · doi:10.1122/8.0000556 · Cited by 28
“where P( ~ D) = ∫ P( ~ D|~ θ)P(~ θ) d~ θ is a normalizing constant [82].”
Kutsukake, N., & Innan, H. (2012) · Evolution · doi:10.1111/j.1558-5646.2012.01775.x · Cited by 33
“…compared simple settings, it will become computationally intensive with increases in the amount of data and the number of parameters In such situations, the ABC inference process should be implemented in the Markov chain Monte Carlo (MCMC) framework (Marjoram et al. 2003; Marjoram and Tavare 2006).”
MethodLopes, J. S., & Boessenkool, S. (2009) · Conservation Genetics · doi:10.1007/s10592-009-0032-9 · Cited by 33
“…approach and gave the technique its name: the study by Beaumont et al. (2002) that explicitly recognised the technique as a method of conditional density estimation and provided refinements based on local linear regression; and the embedding of the ABC method within MCMC by Marjoram et al. (2003).”
MethodWilkinson, D. J. (2011) · Oxford University Press eBooks · doi:10.1093/acprof:oso/9780199694587.003.0023 · Cited by 41
Interrogating theoretical models of neural computation with emergent property inference
Bittner, S. R., Palmigiano, A., Piet, A. T., Duan, C. A., Brody, C. D., Miller, K. D., & Cunningham, J. P. (2021) · eLife · doi:10.7554/elife.56265 · Cited by 32
“Early work used rejection sampling techniques ( Beaumont et al., 2002 ; Marjoram et al., 2003 ; Sisson et al., 2007 ), but EPI and another recently developed methodology ( Gonc¸alves et al., 2019 ) employ deep learning to improve efficiency and provide flexible approximations.”
BackgroundMethodBayesian design strategies for synthetic biology
Barnes, C. P., Silk, D., & Stumpf, M. P. H. (2011) · Interface Focus · doi:10.1098/rsfs.2011.0056 · Cited by 33
DeepDive: estimating global biodiversity patterns through time using deep learning
Cooper, R. B., Flannery‐Sutherland, J. T., & Silvestro, D. (2024) · Nature Communications · doi:10.1038/s41467-024-48434-7 · Cited by 30
Kacprzak, T., Herbel, J., Amara, A., & Refregier, A. R. (2018) · Journal of Cosmology and Astroparticle Physics · doi:10.1088/1475-7516/2018/02/042 · Cited by 27
“It has been implemented within the Monte Carlo Markov Chain (MCMC) framework [1, 13], using Sequential Monte Carlo [14, 15] and Population Monte Carlo [11].”
MethodSun, L., Lee, C., & Hoeting, J. A. (2015) · Environmetrics · doi:10.1002/env.2353 · Cited by 30
“Various ABC algorithms have been proposed, such as rejection-based ABC (Pritchard et al., 1999), regression-based ABC (Beaumont et al., 2002), and ABC Markov chain Monte Carlo (MCMC) (Marjoram et al., 2003).”
BackgroundInfluentialBayesian likelihood-free localisation of a biochemical source using multiple dispersion models
Ristić, B., Gunatilaka, A. H., Gailis, R. M., & Skvortsov, A. T. (2014) · Signal Processing · doi:10.1016/j.sigpro.2014.08.023 · Cited by 27
“been proposed in the single model case, such as the ABC MCMC sampler [21] and a few versions of the ABC sequential Monte Carlo (SMC) sampler [22, 23, 24].”
MethodLikelihood-free simulation-based optimal design with an application to spatial extremes
Hainy, M., Müller, W. G., & Wagner, H. (2015) · Stochastic Environmental Research and Risk Assessment · doi:10.1007/s00477-015-1067-8 · Cited by 27
Quantifying Two-Dimensional Filamentous and Invasive Growth Spatial Patterns in Yeast Colonies
Binder, B. J., Sundstrom, J. F., Gardner, J. M., Jiranek, V., & Oliver, S. G. (2015) · PLoS Computational Biology · doi:10.1371/journal.pcbi.1004070 · Cited by 27
Schuwirth, N., & Reichert, P. (2013) · Ecology · doi:10.1890/12-0591.1 · Cited by 29
“Note that this acceptance–rejection technique is a simple form of a ‘‘likelihood-free’’ (without evaluation of the likelihood function) implementation of Bayesian inference as used in approximate Bayesian computation (e.g., Marjoram et al. 2003).”
MethodAn extended empirical saddlepoint approximation for intractable likelihoods
Fasiolo, M., Wood, S. N., Hartig, F., & Bravington, M. V. (2018) · Electronic Journal of Statistics · doi:10.1214/18-ejs1433 · Cited by 44
“We perform the sampling step using the MCMC approach of Marjoram et al. (2003), and we maximize the approximate posterior using the mean shift algorithm (Fukunaga and Hostetler, 1975).”
MethodAlahmadi, A. A., Flegg, J. A., Cochrane, D. G., Drovandi, C., & Keith, J. M. (2020) · Royal Society Open Science · doi:10.1098/rsos.191315 · Cited by 34
“ABC rejection sampling is very simple to implement, though it can suffer from extremely low acceptance rates when the prior distribution is dissimilar to the posterior distribution [9].”
BackgroundMethodInfluentialWalker, D. M., Allingham, D., Lee, J. H. W., & SMALL, M. P. (2009) · Physica A Statistical Mechanics and its Applications · doi:10.1016/j.physa.2009.09.053 · Cited by 27
“Instead, we turn to a form of MCMC called Approximate Bayesian Computation (ABC) [10] which instead relies on model simulation and realization.”
MethodLikelihoods From Summary Statistics: Recent Divergence Between Species
Leman, S. C., Chen, Y., Stajich, J., Noor, M. A. F., & Uyenoyama, M. K. (2005) · Genetics · doi:10.1534/genetics.104.040402 · Cited by 28
“Likelihoods from summary statistics: Marjoram et al. (2003) introduced a Markov chain Monte Carlo (MCMC) method that directly approximates the posterior distribution of the model, obviating the need to determine the probability of the data.”
BackgroundMethodTime, Species, and Separating Their Effects on Trait Variance in Clades
Bokma, F. (2010) · Systematic Biology · doi:10.1093/sysbio/syq029 · Cited by 27
“A recent technique to estimate parameters without a likelihood function is approximate Bayesian computation (ABC; Marjoram et al. 2003).”
MethodSimulation framework for generating intratumor heterogeneity patterns in a cancer cell population
Iwasaki, W. M., & Innan, H. (2017) · PLoS ONE · doi:10.1371/journal.pone.0184229 · Cited by 32
“To do so, we aimed to ensure that simulations do not take a very long time so that it can be used within the framework of simulation-based inference as outlined in Marjoram et al. [22] (see also refs therein).”
MethodApproximating Bayes in the 21st Century
Martin, G. M., Frazier, D. T., & Robert, C. P. (2023) · Statistical Science · doi:10.1214/22-sts875 · Cited by 44
“One such approach is to insert an MCMC step, and associated proposal distribution, within Algorithm 2 in order to more effectively explore the space, which yields an ABC-MCMC algorithm (Marjoram et al., 2003).”
MethodInfluentialScalable Bayesian Inference for the Inverse Temperature of a Hidden Potts Model
Moores, M. T., Nicholls, G. K., Pettitt, A. N., & Mengersen, K. L. (2018) · Bayesian Analysis · doi:10.1214/18-ba1130 · Cited by 36
“Instead we have based Algorithm 4 on the ABC-MCMC algorithm of Marjoram et al. [2003].”
MethodImproving Approximate Bayesian Computation via Quasi-Monte Carlo
Buchholz, A., & Chopin, N. (2018) · Journal of Computational and Graphical Statistics · doi:10.1080/10618600.2018.1497511 · Cited by 26
“For instance the use of MCMC as by Marjoram et al. (2003) has improved the simulation of ABC posterior samples over the simple accept–reject algorithm.”
MethodBayesian Probabilistic Power Flow Analysis Using Jacobian Approximate Bayesian Computation
Zuluaga, C. D., & Álvarez, M. A. (2018) · IEEE Transactions on Power Systems · doi:10.1109/tpwrs.2018.2810641 · Cited by 26
“lem is addressed by an ABC extension presented in [15], which is based on Markov Chain Monte Carlo (MCMC) [21].”
MethodLewis, J. R., MacEachern, S. N., & Lee, Y. (2021) · Bayesian Analysis · doi:10.1214/21-ba1257 · Cited by 43
“…posterior approximation methods which has recently experienced success in applications to epidemiology, genetics, and quality control (see, for example, Tavaré et al., 1997; Pritchard et al., 1999; Beaumont et al., 2002; Marjoram et al., 2003; Fearnhead and Prangle, 2012; Drovandi et al., 2015).”
MethodImplicit Estimation of Ecological Model Parameters
Weir, B., Miller, R. N., & Spitz, Y. H. (2013) · Bulletin of Mathematical Biology · doi:10.1007/s11538-012-9801-6 · Cited by 26
Searle, K. R., O’Brien, S. H., Jones, E. L., Cook, A. S. C. P., Trinder, M., Mcgregor, R., Donovan, C., McCluskie, A., Daunt, F., & Butler, A. (2023) · ICES Journal of Marine Science · doi:10.1093/icesjms/fsad025 · Cited by 25
Likelihood-Free Inference in High-Dimensional Models
Kousathanas, A., Leuenberger, C., Helfer, J., Quinodoz, M., Foll, M., & Wegmann, D. (2016) · Genetics · doi:10.1534/genetics.116.187567 · Cited by 25
“The original ABC-MCMC algorithm proposed by Marjoram et al. (2003) [3] is: 1.”
MethodWhite, S. R., Kypraios, T., & Preston, S. (2013) · Statistics and Computing · doi:10.1007/s11222-013-9432-2 · Cited by 30
“Third, we consider an integer-valued time series model called INAR(1), a model for which the likelihood is available (albeit awk-ward to compute) and enables comparison of our approach with a “gold standard” MCMC approach.”
MethodInfluentialChoosing the Summary Statistics and the Acceptance Rate in Approximate Bayesian Computation
Blum, M. G. B. (2010) · Physica-Verlag eBooks · doi:10.1007/978-3-7908-2604-3_4 · Cited by 25
“Approximate Bayesian Computation (ABC) encompasses a family of likelihood free algorithms for performing Bayesian inference (Beaumont et al. (2002), Marjoram et al. (2003)).”
Social Networks in Policy Making
Battaglini, M., & Patacchini, E. (2019) · Annual Review of Economics · doi:10.1146/annurev-economics-080218-030419 · Cited by 31
Likelihood-free parallel tempering
Baragatti, M., Grimaud, A., & Pommeret, D. (2012) · Statistics and Computing · doi:10.1007/s11222-012-9328-6 · Cited by 29
“Algorithms have been proposed by Marjoram et al. (2003), Bortot et al. (2007) and Ratmann et al. (2007), the reference method being the one of Marjoram et al. (2003) who suggested the use of a Markov kernel q(.”
MethodWu, Q., Smith‐Miles, K., & Tian, T. (2014) · BMC Bioinformatics · doi:10.1186/1471-2105-15-s12-s3 · Cited by 26
“For the case of parameter estimation when likelihoods are analytically or computationally intractable, approximate Bayesian computation (ABC) methods have been applied successfully [20,21].”
MethodScalable inference for Markov processes with intractable likelihoods
Owen, J., Wilkinson, D. J., & Gillespie, C. S. (2014) · Statistics and Computing · doi:10.1007/s11222-014-9524-7 · Cited by 25
“to MCMC schemes which avoid calculation of the likelihood, as well as sequential Monte Carlo algorithms, which typically exhibit better acceptance rates than the simple rejection sampler (Marjoram et al. 2003; Del Moral et al. 2006; Sisson et al. 2007; Toni et al. 2009).”
BackgroundMethodWhen virulence originates from non-agricultural hosts: New insights into plant breeding
Leroy, T., Cam, B. L., & Lemaire, C. (2014) · Infection Genetics and Evolution · doi:10.1016/j.meegid.2013.12.022 · Cited by 25
“Distinguishing between the two scenarios is also possible using likelihood-free inference methods, often referred to as Approximate Bayesian Computation (ABC).”
BackgroundInfluentialLikelihood free inference for Markov processes: a comparison
Owen, J., Wilkinson, D. J., & Gillespie, C. S. (2015) · Statistical Applications in Genetics and Molecular Biology · doi:10.1515/sagmb-2014-0072 · Cited by 24
“Optimal choices for N were subject to interest in Pitt et al. (2012) and Doucet et al.”
BackgroundMonte Carlo Strategies for Selecting Parameter Values in Simulation Experiments
Leigh, J., & Bryant, D. (2015) · Systematic Biology · doi:10.1093/sysbio/syv030 · Cited by 26
“Our method for sampling from P( |R) is based on the MCMC without likelihoods algorithm of Marjoram et al. (2003). Their algorithm takes a data set D and generates samples from the posterior density P( |D).”
InfluentialHua, X., & Bromham, L. D. (2015) · Methods in Ecology and Evolution · doi:10.1111/2041-210x.12531 · Cited by 25
“Alternatively, our backward tree simulation allows efficient implementation of an approximated Bayesian computation approach (Marjoram et al. 2003).”
MethodTangled Trees: The Challenge of Inferring Species Trees from Coalescent and Noncoalescent Genes
Anderson, C. N. K., Liu, L., Pearl, D. K., & Edwards, S. V. (2012) · Methods in molecular biology · doi:10.1007/978-1-61779-585-5_1 · Cited by 27
Zhang, J., Dennis, T. E., Landers, T. J., Bell, E. A., & Perry, G. L. W. (2017) · Ecological Modelling · doi:10.1016/j.ecolmodel.2017.07.017 · Cited by 24
Metadata
- Authors
- Marjoram, P., Molitor, J., Plagnol, V., & Tavaré, S.
- Affiliations
- University of Southern California (US)
- Venue
- Proceedings of the National Academy of Sciences
- Publisher
- National Academy of Sciences
- Published
- 2003-12-08
- Volume
- 100
- Issue
- 26
- Pages
- 15324-15328
- ISSN
- 0027-8424 (ISSN-L)
- 1091-6490
- PMC
- PMC307566
- OpenAlex
- W2045973738
- Cited by
- 1,199
- References
- 20
- Citations by year
- 2026: 26 · 2025: 38 · 2024: 54 · 2023: 54 · 2022: 57 · 2021: 72 · 2020: 83 · 2019: 79 · 2018: 88 · 2017: 68 · 2016: 76 · 2015: 79 · 2014: 86 · 2013: 64 · 2012: 60
- FWCI
- 12.01
- Type
- journal article
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Locations
- Proceedings of the National Academy of Sciences (publishedVersion)
- PubMed Central (publishedVersion)
- PubMed (publishedVersion)
- UCL Discovery (University College London) (submittedVersion)
- Funders
- National Institutes of Health (5r01gm058897-03, GM58897)
- National Institute of General Medical Sciences (GM 58897)
- Division of Mathematical Sciences
- Topics
- Physical Sciences › Mathematics › Statistics and Probability › Markov Chains and Monte Carlo Methods
- Physical Sciences › Mathematics › Statistics and Probability › Statistical Methods and Bayesian Inference
- Life Sciences › Biochemistry, Genetics and Molecular Biology › Genetics › Genetic diversity and population structure
- Keywords
- Markov chain Monte Carlo, simulation-based inference, posterior sampling, maximum likelihood estimation, population genetics
- MeSH terms
- Algorithms, Computer Simulation, DNA, DNA, Mitochondrial, Biological Evolution, Genetics, Population, Humans, Markov Chains, Models, Biological, Monte Carlo Method, Stochastic Processes, Likelihood Functions
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.