Abstract

The current theoretical frameworks of biochemical computations mainly reflect Turing-like information processing with stable states. Although a broad spectrum of information processing problems can be treated within this framework, it is still challenging to understand how this leads to balancing between opposed features such as robustness and plasticity, as well as biochemical computations in real-time. We have identified that organization at criticality on the one hand enhances computational features of systems where the information is stored in the stable biochemical states, and on the other hand, serves as a basis of computation with meta-stable states, thereby providing real-time information processing. We substantiate these findings through two experimental examples: a developmental problem to address how robust proportioning between differentiated cell fates emerges, and an example of cellular sensing of and responsiveness to non-stationary growth factor patterns.