Engineering microbial consortia for distributed signal processing

Area of Science:

• Synthetic biology
• Microbial ecology
• Computational biology

Background:

• Inferring input signals from biological readouts is crucial.
• Engineered biosensors typically require laborious orthogonalization to prevent crosstalk.

Purpose of the Study:

• To develop a method for inferring multiplexed concentrations without eliminating sensor crosstalk.
• To leverage microbial communities for distributed sensing and decode collective responses.

Main Methods:

• Distributing sensing across a microbial community.
• Decoding time-resolved collective responses.
• Coupling kinetic modeling with machine learning for input concentration mapping.

Main Results:

• Quantitative inference of multiplexed concentrations was achieved despite sensor crosstalk and indirect responses.
• The method was validated in diverse scenarios, including antibiotic combinations and environmental samples.
• Demonstrated successful inference in communities with varying degrees of crosstalk.

Conclusions:

• Distributed dynamic sensing enables multiplexed measurements by tolerating non-orthogonal responses.
• This approach broadens the applicability of biological systems for complex input detection.
• Reproducible and distinguishable community response trajectories can disambiguate input combinations.