This thesis focuses on the development of new techniques for modelling, parameterising, and analysing fluorescent proteins. Computational studies of fluorescent proteins are plagued by a steep learning curve, external tools demanding differing inputs, and silent errors that go unnoticed for months. Here we present tools aimed at making computational chemistry more accessible to fluorescent protein studies, together with two studies that demonstrate their potential. Chapter 4 introduces a Graphical User Interface aimed at abstracting command-line tools and other processes necessary for the creation of files ready for computation. Each of the (up to thirteen) stages is presented as an interactable object, relaying information and errors to the user. Relevant decisions and actions are exposed, while those disconnected from the question at hand such as file format conversion are performed under-the-hood. Chapter 5 extends the software with the capability to simulate saturation mutagenesis. Two methods for the optimisation of amino acid sidechains are described and compared. The sensitivity of the Green Fluorescent Protein chromophore toward mutation of its local environment is investigated. Molecular orbital and electron density analysis provides an explanation of the robustness of the chromophore. Chapter 6 is centred on the situation where multiple feasible reaction pathways between a set of reactants and products exist. While this situation has been covered extensively in literature for two coordinates, no examples of a true three-coordinate potential energy surface have been found. Interpolated scans are used to produce three-dimensional grids which are visualised using our own shader code. This technique is applied to the guanine-cytosine DNA base pair, followed by the proton transfer that links the I and A states of Green Fluorescent Protein. By comparing the electronic states along the reaction profile, an explanation of the increased reactivity on the excited state is given.