Model selection and parameter estimation for root architecture models using likelihood-free inference.

Clare Ziegler,Iain G Johnston,Rosemary J Dyson

doi:10.1098/rsif.2019.0293

Abstract

Plant root systems play vital roles in the biosphere, environment and agriculture, but the quantitative principles governing their growth and architecture remain poorly understood. The ‘forward problem’ of what root forms can arise from given models and parameters has been well studied through modelling and simulation, but comparatively little attention has been given to the ‘inverse problem’: what models and parameters are responsible for producing an experimentally observed root system? Here, we propose the use of approximate Bayesian computation (ABC) to infer mechanistic parameters governing root growth and architecture, allowing us to learn and quantify uncertainty in parameters and model structures using observed root architectures. We demonstrate the use of this platform on synthetic and experimental root data and show how it may be used to identify growth mechanisms and characterize growth parameters in different mutants. Our highly adaptable framework can be used to gain mechanistic insight into the generation of observed root system architectures.

Highlights

Root systems are essential to plants’ structure and uptake of water and nutrients and constitute more than 5% by mass of the total global carbon budget [1]
Branches elongate according to the primary root growth law multiplicatively scaled
There is a very little constraint in the value of lmax, suggesting little reliance on the value, smaller values appear to be favoured for the friendly phenotype. These results demonstrate that the physical parameters governing root architecture growth can be learned using this approximate Bayesian computation (ABC) sequential Monte Carlo (SMC) approach, and uncertainty in these learned outcomes quantified

Summary

Introduction

Root systems are essential to plants’ structure and uptake of water and nutrients and constitute more than 5% by mass of the total global carbon budget [1]. The shape of a plant’s root system is generated by a variety of physiological and signalling pathways within the plant, and understanding the generation of this system opens paths to its optimization to maximize crop yield [6]. Despite this importance, the mechanisms underlying root growth remain challenging to quantitatively understand [7,8,9]. More recent advances have facilitated the imaging of plants in situ through the use of X-ray m-computed tomography [10], magnetic resonance imaging scanning [11,12] and transparent soil [13], which have been used to investigate root soil exploration and uptake of water and nutrients

Methods

Results

Conclusion