Porous organic molecules have internal pores readily occupied by gases, solvent or other guests. These molecules can form porous molecular materials with possible application in storage and separation. In this thesis, the properties of isolated porous organic molecules are used as a proxy to the bulk or in-solution applications. As a result, software was first developed for the automated and precise structural characterisation of porous organic molecules. This allows one to easily calculate window diameters to study the thermal window size fluctuations and predict guest diffusion in the bulk. A screening of previously reported porous organic molecules for the application of Xe/Kr separation allowed the most promising material, Noria, to be identified. This was possible with a combination of molecular modelling, electronic structure calculations and structural analysis using the developed software. The experimental Xe/Kr selectivity of Noria, not previously considered for this application, was shown to be comparable with the best performing porous materials. Next, eight synthetically realised porous organic cages were studied as possible C60 fullerene encapsulants. The relative orientation of the C60 fullerene in the pore was shown to have little to no effect on the binding energy and the encapsulation of the C60 during cage formation was determined as the likely mechanism of encapsulation. Lastly, a function-led material design approach was developed. An evolutionary algorithm was used to generate possible C60 encapsulants from a database of precursors. The resulting porous organic cages are structurally similar to some recently synthetically realised cages and some found in the literature. In summary, presented in this thesis is software, a methodology and results that can further advance the computational function-led materials discovery for specific applications.