In existing models, the interaction energy is proportional to the interaction area between cells, which cannot describe the direct measurements obtained with OT. First, considering the cell interaction in protein solution, the “depletion layer” model is not applicable to describe it, as the depletion forces exclude the possibility of cell interaction without their spontaneous overlapping. On the other hand, the “bridging” model can be still used, assuming that the membranes of cells cannot be brought close enough to each other to form a bridge if there is no external force. Thus, the cells interact strongly when they are brought intact with OT. However, the ordinary “bridging” models describe the interaction such that bridges are uniformly distributed,35 while our experimental results show that the interaction becomes stronger as cells are gradually moved to the separation point. We propose to extend the “bridging” model with a possibility of bridge mobility, which fits excellently with experimental results. If the bridges can move along with cell separation and also form new ones as the cells are moved from each other, as schematically illustrated in Fig. 7, the total interaction force will be increased. A similar approach was first introduced by Evans and Leung48 to describe the agglutination interaction between RBCs induced by white germ agglutinin. It is necessary to mention that the agglutination interaction was about 2 orders stronger than one observed in our work. Further, the model was developed by Tozeren et al.49 based on the interaction between a T-killer cell and its target. This model can be considered an extended “cross-bridge” model of the cell interaction, and, as illustrated schematically in Fig. 7, it assumes that binding of cells is possible via cross-bridges. These bridges are (1) formed at the initial interaction area, and (2) as cells are moved from each other, these bridges migrate along with the cell membrane (or break apart) and new bridges are formed; (3) the total number of bridges increases and so does the interaction force between the cells if migration and formation of bridges prevails over breaking of bridges. This model can also explain why the increase of interaction over cell separation is not observed in the measurements performed with whole blood aggregometry methods. In particular, in the device based on the microfluidic chamber (RheoScan), initially high and gradually decreasing shear stress is applied, which could mean that the bridges are not allowed to migrate.