Mechanisms enabling complex multi-class biochemical classification

Determine the biochemical mechanisms by which cellular systems such as glycosylation networks or adaptive immune recognition achieve complex multi-class classification among large numbers of possibilities, potentially via microscopic sensing events such as estimation of antigen binding affinities.

Background

The paper frames many cellular decision-making processes as classification tasks implemented by non-equilibrium biochemical networks. While the authors quantify expressivity limits and how input promiscuity can increase capacity, they note that certain biological systems perform remarkably complex multi-class decisions (e.g., hundreds of glycan states or immune antigen recognition).

The explicit open question concerns the underlying biochemical strategies by which such systems accomplish these high-capacity classifications. The context suggests mechanisms could involve microscopic sensing or affinity estimation, but a definitive characterization remains unresolved.

References

How these biochemical systems achieve these complex classification tasks (e.g., through microscopic sensing events like estimating antigen binding affinities) remains an important and open question.

— Limits on the computational expressivity of non-equilibrium biophysical processes  (2409.05827 - Floyd et al., 2024) in Section: Storing more classes by increasing input promiscuity (Main text)