Computational hypothesis generation with genome-side metabolic reconstructions: in-silico prediction of metabolic changes in the freshwater model organism Daphnia to environmental stressors

Bradbury, James (2018). Computational hypothesis generation with genome-side metabolic reconstructions: in-silico prediction of metabolic changes in the freshwater model organism Daphnia to environmental stressors. University of Birmingham. Ph.D.

Preview

Bradbury18PhD.pdf
PDF - Accepted Version
Download (6MB)

Abstract

Computational toxicology is an emerging, multidisciplinary field that uses in-silico modelling techniques to predict and understand how biological organisms interact with pollutants and environmental stressors. Genome-wide metabolic reconstruction (GWMR) is an in-silico modelling technique that aims to represent the metabolic capabilities of an organism.

Daphnia is an emerging model species for environmental omics whose underlying biology is still being uncovered. Creating a metabolic reconstruction of Daphnia and applying it in an environmental computational toxicology setting has the potential to aid in understanding its interaction with environmental stressors. Here, the fist GWMR of D. magna is presented, which is built using METRONOME, a newly developed tool for automated GWMR of new genome sequences. Active module identification allows for omics data sets to be integrated into in-silico models and uses optimisation algorithms to find hot-spots within networks that represent areas that are significantly impacted based on a toxicogenomic transcriptomics dataset. Here, a method that uses the active modules approach in a predictive capacity for computational hypothesis generation is introduced to predict unknown metabolic responses to environmentally relevant human-induced stressors.

A computational workflow is presented that takes a new genome sequence, builds a GWMR and integrates gene expression data to make predictions of metabolic effects. The aim is to introduce an element of hypothesis generation into the untargeted metabolomics experimental workflow. A study to validate this approach using D. magna as the target organism is presented, which uses untargeted Liquid-Chromatography Mass Spectrometry (LC-MS) to make metabolomics measurements. A software tool MUSCLE is presented that uses multi-objective closed-loop evolutionary optimisation to automatically develop LC-MS instrument methods and is used here to develop the analytical method.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):