Why E. coli encodes far less information than molecule counting allows
In plain words
The gut bacterium E. coli swims toward food by sensing chemical gradients, and it climbs them nearly as fast as the information it collects allows. Yet it collects about a hundred times less information than the random arrival of molecules at its receptors would permit, and the source of this loss is unknown.
Precise statement
For E. coli in shallow gradients of methyl-aspartate, the measured information rate from ligand input to chemotaxis signaling (kinase) activity is about $10^{-2}$ of the molecule-arrival (Berg-Purcell type) limit, while drift speed is near the bound set by the measured information rate (Mattingly and co-workers, 2021 and 2026). Identify which internal noise source (receptor-cluster switching, CheR/CheB methylation noise, CheY phosphorylation noise) sets the information rate, and whether it reflects a physical constraint or a trade-off; an answer is a quantitative noise budget that reproduces the measured information rate.
What would settle it
Single-cell measurements of each candidate noise source and of the information rate in the same strains, with a model that closes the information budget and predicts the effect of mutations that change each source.
Status in the literature
Unverified note
Single-cell measurements in 2026 placed E. coli two orders of magnitude below the molecule-arrival limit and attributed the loss to internal signaling noise (Nature Physics, doi 10.1038/s41567-025-03111-4); the dominant source is not identified.