Pakman: a Modular, Efficient and Portable Tool for Approximate Bayesian Inference

The development of high-throughput techniques in the biological sciences has resulted in an abundance of experimental data. At the same time, mathematical models are becoming increasingly popular in the biological sciences. Combined, these parallel advances have the potential to greatly expand our understanding of biological processes through mathematical modelling and data-driven parameter inference. When the data are noisy or the mathematical model is inherently stochastic, we can apply Bayesian methods for parameter inference and model selection. Moreover, even when the likelihood function is unknown or intractable, it is still possible to make progress by applying a method known as approximate Bayesian computation (ABC) (Marjoram, Molitor, Plagnol, & Tavare, 2003; Toni, Welch, Strelkowa, Ipsen, & Stumpf, 2009).