Extended Data Fig. 10: Parallels and differences between pattern recognition in our synthetic approach and in potential biological systems. | Nature

Extended Data Fig. 10: Parallels and differences between pattern recognition in our synthetic approach and in potential biological systems.

From: Pattern recognition in the nucleation kinetics of non-equilibrium self-assembly

Extended Data Fig. 10

While we studied pattern recognition using a specific set of molecules (DNA) in an in vitro system, the concepts behind our work have potential relevance to biological systems built out of molecules of different nature and in different conditions. a, Pattern recognition was triggered in our system by lowering temperature over time which drives the self-assembly process forward. Instead, in the cell, concentrations of molecular components can rise over time (e.g., through gene expression), leading to nucleation and self-assembly and thus pattern recognition. b, Timescale of pattern recognition is controlled by system-specific aspects and also general trends. System specific: DNA-specific processes such as tile attachment, detachment and restructuring set the timescale of nucleation and growth. Living systems can use active mechanisms to control nucleation timescales in addition to the concentration effects described here. Broader trends: our theoretical work supports a general relationship between the speed of pattern recognition (e.g., by working at a lower temperature), the size of critical nuclei and thus the complexity of pattern recognition. c, The winner-take-all effect in our work enhanced selectivity by exploiting the depletion of shared components. Biomolecular systems, such as macromolecular complexes56 and multicomponent phase condensates46 are thought to share components as well, potentially enabling a winner-take-all effect in cells. d, In the biological context, the inputs could represent physiological or environmental signals encoded in the relative concentrations of many species of molecules. Some patterns of enhanced concentration may not lead to self-assembly or phase condensation if those components with enhanced concentration are not colocalized on a structure or reinforce a nucleation pathway for a condensate; but an alternative pattern of high concentrations could lead to assembly of one of several assemblies or condensates. e, Such sensitivity of kinetic pathways to concentration patterns can be exploited for complex decision-making in numerous aspects of cellular physiology, or may provide compact and robust control mechanisms for cell-scale molecular robots. See also Supplementary Information section 1.5 for how a pixel-to-tile map could be physically incarnated.

Back to article page