Computer simulation studies aimed at elucidating the phase behavior of crude oils inevitably require atomistically-detailed models of representative molecules. For the lighter fractions of crudes, such molecules are readily available, as the chemical composition can be resolved experimentally. Heavier fractions pose a challenge, on one hand due to their polydispersity and on the other due to poor description of the morphology of the molecules involved. The Quantitative Molecular Representation (QMR) approach is used here to generate a catalogue of 100 plausible asphaltene and resin structures based on elemental analysis and 1 H-13 C NMR spectroscopy experimental data. The computer-generated models are compared in the context of a review of previously proposed literature structures and categorized by employing their molecular weights, double bond equivalents (DBE) and hydrogen to carbon (H/C) ratios. Sample atomistic molecular dynamics simulations were carried out for two of the proposed asphaltene structures with contrasting morphologies, one island-type and one archipelago-type, at 7 wt% in either toluene or heptane. Both asphaltene models, which shared many characteristics in terms of average molecular weight, chemical composition and solubility parameters showed marked differences in their aggregation behavior. The example showcases the importance of considering diversity and polydispersity when considering molecular models of heavy fractions.