We present a novel family of deep neural architectures, named partially exchangeable networks (PENs) that leverage probabilistic symmetries. By design, PENs are invariant to block-switch transformations, which characterize the partial exchangeability properties of conditionally Markovian processes. Moreover, we show that any block-switch invariant function has a PEN-like representation. The DeepSets architecture is a special case of PEN and we can therefore also target fully exchangeable data. We employ PENs to learn summary statistics in approximate Bayesian computation (ABC). When comparing PENs to previous deep learning methods for learning summary statistics, our results are highly competitive, both considering time series and static models. Indeed, PENs provide more reliable posterior samples even when using less training data.
We introduce the sequential neural posterior and likelihood approximation (SNPLA) algorithm. SNPLA is a normalizing flows-based algorithm for inference in implicit models. Thus, SNPLA is a simulation-based inference method that only requires simulations from a generative model. Compared to similar methods, the main advantage of SNPLA is that our method jointly learns both the posterior and the likelihood. SNPLA completely avoid Markov chain Monte Carlo sampling and correction-steps of the parameter proposal function that are introduced in similar methods, but that can be numerically unstable or restrictive. Over four experiments, we show that SNPLA performs competitively when utilizing the same number of model simulations as used in other methods, even though the inference problem for SNPLA is more complex due to the joint learning of posterior and likelihood function.
Mathematical modelling is an invaluable tool to describe dynamic cellular processes and to rationalise cell-to-cell variability within the population. This requires statistical methods to infer unknown model parameters from dynamic, multi-individual data accounting for heterogeneity caused by both intrinsic and extrinsic noise. Here we present PEPSDI, a scalable and flexible framework for Bayesian inference in state-space mixed-effects stochastic dynamic single-cell models. Unlike previous frameworks, PEPSDI imposes a few modelling assumptions when inferring unknown model parameters from time-lapse data. Specifically, it can infer model parameters when intrinsic noise is modelled by either exact or approximate stochastic simulators, and when extrinsic noise is modelled by either time-varying, or time-constant parameters that vary between cells. This allowed us to identify hexokinase activity as a source of extrinsic noise, and to deduce that sugar availability dictates cell-to-cell variability in the budding yeast Saccharomyces cerevisiae SNF1 nutrient sensing pathway.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.