Bayesian uncertainty analysis establishes the link between the parameter space of a complex model of hormonal crosstalk in Arabidopsis root development and experimental measure

by   Samuel E. Jackson, et al.

A major challenge in plant developmental biology is to understand how plant growth is coordinated by interacting hormones and genes. To meet this challenge, it is important to formulate a mathematical model. For the mathematical model to best describe the true biological system, it is necessary to understand the parameter space of the model. We develop sequential history matching methodology, using Bayesian emulation, to comprehensively explore the link between subsets of experimental measurements and the model parameter space of a complex hormonal crosstalk model for Arabidopsis root growth. Using 22 trends, the set of acceptable inputs is reduced to 6.1 × 10^-5% of the original space. It is revealed that 5 further trends, for exogenous application of ACC, facilitate an additional reduction of 3 orders of magnitude. Moreover, the final 5 trends, for measurement of the POLARIS gene expression, refocused the set by another 2 orders of magnitude. This indicates that both types of experiments (namely exogenous application of a hormone and measurement of the gene expression) can significantly reduce the set of acceptable parameters. Hence, we provide insight into the constraints placed upon the model structure by, and the biological consequences of, measuring subsets of observations. In particular, we learn about the joint structure of output constraints on the input space. An important example is how different regulatory relationships, such as negative or positive regulation of auxin biosynthesis by cytokinin, affect constraints on the non-implausible parameter space. The link between parameter space and experimental measurements in a complex hormonal crosstalk model for Arabidopsis root growth is established using Bayesian uncertainty analysis. Our approach establishes a novel methodology for understanding the link between experiments, model and parameter space.


page 1

page 13

page 16

page 17

page 26

page 27

page 28

page 30


Latent Projection BNNs: Avoiding weight-space pathologies by learning latent representations of neural network weights

While modern neural networks are making remarkable gains in terms of pre...

Reducing Parameter Space for Neural Network Training

For neural networks (NNs) with rectified linear unit (ReLU) or binary ac...

Dissecting graph measure performance for node clustering in LFR parameter space

Graph measures that express closeness or distance between nodes can be e...

Analyzing Inverse Problems with Invertible Neural Networks

In many tasks, in particular in natural science, the goal is to determin...

Identifying the Parametric Occurrence of Multiple Steady States for some Biological Networks

We consider a problem from biological network analysis of determining re...

Emulation and History Matching using the hmer Package

Modelling complex real-world situations such as infectious diseases, geo...

Please sign up or login with your details

Forgot password? Click here to reset