Model-Driven Data Collection for Biological Systems

TL;DR

This paper introduces an information-theoretic sequential experimental design method for biological systems, optimizing sampling time and observable choice to efficiently estimate unknown parameters under practical constraints.

Contribution

It presents a novel adaptive design approach based on mutual information to improve parameter estimation efficiency in biological experiments.

Findings

Optimal sampling time and observable can be determined via information-theoretic sensitivity analysis.

The proposed method reduces uncertainty faster than non-adaptive designs.

Application to two biological systems demonstrates effectiveness.

Abstract

For biological experiments aiming at calibrating models with unknown parameters, a good experimental design is crucial, especially for those subject to various constraints, such as financial limitations, time consumption and physical practicability. In this paper, we discuss a sequential experimental design based on information theory for parameter estimation and apply it to two biological systems. Two specific issues are addressed in the proposed applications, namely the determination of the optimal sampling time and the optimal choice of observable. The optimal design, either sampling time or observable, is achieved by an information-theoretic sensitivity analysis. It is shown that this is equivalent with maximizing the mutual information and contrasted with non-adaptive designs, this information theoretic strategy provides the fastest reduction of uncertainty.