The design and discovery of potential novel materials is critical for the advancement of climate change mitigation technologies. In this respect, metal-organic frameworks (MOFs) have received considerable attention over the last two decades. The combination of high surface areas, wide range of pores, and tunable chemistry makes them the 'designer's' choice of materials for energy-related, medicine-related, and other applications. However, the chemical space of MOFs is so enormous (with trillions of possible structures) that finding the optimal material for a given application has become a very challenging problem and is often referred to as the 'holy grail of material science'. This thesis attempts to address this problem by focusing on a diversity-driven discovery of MOFs for different energy-related applications such as carbon capture, hydrogen storage, methane separation, and photocatalysis. Using a combination of computational material design, molecular simulation, and machine learning techniques, our search methodology helps us to effectively explore the MOF chemical space and identify the most interesting regions of this space for the respective applications. We find that through our approach we are able to tailor-make MOF structures that compete strongly against the top-performing ones reported in literature for the each of the studied application. We also take a step further and study the structures though different levels of simulation: starting from their atomic constituents right to its performance in actual industrial scenario. Overall, this thesis sheds light and provides directions on advancing material design and discovery for energy-related applications.