One of the main goals of computational methods is to identify reasonable geometries for target materials. Organometallic complexes have been investigated in this dissertation research, entailing a significant challenge based on transition metal diversity and the associated complexity of the ligands. A large variety of theoretical methods have been employed to determine ground state geometries of organometallic species. An impressive number of transition metals entailing diverse isomers (e.g., geometric, spin, structural and coordination), different coordination numbers, oxidation states and various numbers of electrons in d orbitals have been studied. Moreover, ligands that are single, double or triple bonded to the transition metal, exhibiting diverse electronic and steric effects, have been investigated. In this research, a novel de novo scheme for structural prediction of transition metal complexes was developed, tested and shown to be successful.