An autonomous system for multi-objective continuous evolution at scale

Area of Science:

• Evolutionary biology
• Synthetic biology
• Biotechnology

Background:

• Natural evolution is high-dimensional, with organisms adapting to multiple pressures simultaneously.
• Directed evolution methods often simplify this landscape to a single selection axis, limiting insights and potential engineering.
• Phage-assisted continuous evolution (PACE) offers multivariate selection but faces scalability challenges due to individual culture requirements.

Purpose of the Study:

• To develop a scalable platform for multivariate directed evolution.
• To enable high-throughput evolutionary tracking and data generation for complex fitness landscapes.
• To overcome the limitations of previous PACE implementations regarding automation and parallelization.

Main Methods:

• Introduction of TurboPRANCE, an open-source robotic platform integrating independently controlled turbidostats with parallel PACE lagoons.
• Implementation of closed-loop control, automated media formulation, programmable dosing, and adaptive scheduling for sustained, asynchronous operation.
• Integration of Nanopore long-read sequencing with DeepVariant for population-level variant tracking and high-resolution evolutionary trajectory analysis.

Main Results:

• TurboPRANCE enables multiplexed selection across diverse regimes with combinatorial pressures and unbounded experimental designs.
• The system maintains cultures at consistent, infectable densities despite unpredictable workflow demands.
• High-resolution, time-resolved evolutionary trajectories and large parallel datasets are generated, suitable for training machine learning models.

Conclusions:

• TurboPRANCE significantly advances the scalability and throughput of multivariate directed evolution.
• The platform facilitates the mapping and engineering of complex fitness landscapes at an unprecedented scale.
• This work provides a powerful tool for fundamental evolutionary studies and applied biotechnology.