The properties of organs, tissues, organoids, and other systems of cells, are influenced by the spatial localization and distribution of their elements. Here, we used networks to describe distributions of cells on a surface where the small-world coefficient (SW) of the networks was varied between SW~1 (random uniform distributions) and SW~10 (clustered distributions). The small-world coefficient is a topological measure of graphs:​ networks withSW > 1 are topologically biased to transmit information. For each system configuration, we then determined the total energy U as the sum of the energies that describe cell-cell interactions – approximated by a harmonic potential. The graph of energy (U) across the configuration space of the networks (SW) is the energy landscape:​ it indicates which configuration a system of cells will likely assume over time. We found that, depending on the model parameters, the energy landscapes of 2D distributions of cells may be of different types:​ from type I to type IV. Type I and type II systems have high probability to evolve into random distributions. Type III and type IV systems have a higher probability to form clustered architectures. A great many of simulations indicated that cultures of cells with high initial density and limited sensing range could evolve into clustered configurations with enhanced topological characteristics. Moreover, the strongest the binding between cells, the greater the likelihood that they will assume configurations characterized by finite values of SW. Results of the work are relevant for those working the field of tissue engineering, regenerative medicine, the formation of in-vitro-models, the analysis of neuro-degenerative diseases.