Systematic Analysis and Identification of Substrates Secreted by Gram-negative Bacteria

Gram-negative bacteria have evolved a wide range of secretion systems as their weapons to export substrate proteins into the surrounding milieu or adjacent target cells. These secreted proteins play vital roles in the struggle against stressful environments, and contribute toward bacterial pathogenesis and their competitive survival in bacterial populations. With the purpose of facilitating statistical analysis and computational prediction of various types of substrate, this thesis aimed to develop a series of analytical and predictive toolkits based on machine learning with the intention to interlink them as an integrative platform and pipeline. Through providing seamless operations between laboratory-confirmed substrates, potential substrate prediction and their inter-relationship analysis, this streamlined tool suite is expected to provide insights into the known substrates and facilitate new substrate discoveries.