Integration of Quality By Design, Programming Languages, and Cross-Platform Automation for Closing the Design, Build, Test Cycle in Metabolic Engineering
Building robust manufacturing processes by metabolic engineering is a task that is highly complex and requires sophisticated tools to design and describe processes, inputs and measurements; as well as to administrate management of knowledge, data, and materials.
We argue that for bioengineering to succeed in the most difficult cases, it will require easier and more routine application of engineering and quality approaches such as design for manufacturing (DfM) and quality by design (QbD) as well as the application of statistically designed experiments to derive detailed empirical models of underlying genetic systems in context with changing environmental conditions. This requires execution of large-scale structured experimentation for which laboratory automation is necessary.
This will be greatly facilitated by the development of expressive, high-level languages that allow reusability of protocols, characterization of their reliability, breaking the artificial separation between design and execution and abstraction of implementation details in favor of functional properties.
To directly address these issues, we have developed Antha, a high-level language (antha-lang) and operating system (anthaOS) for describing, executing and transferring bio-R&D and biomanufacturing processes.
With a particular emphasis on automation, workflows are written in a high-level form avoiding device-specific low-level instructions. The structure is designed to enforce good quality practices, reusability and faciliation of mathematical approaches. This eases transferability between equivalent but distinct laboratory and manufacturing infrastructures and presents an opportunity to increase insight by facilitating more structured and sophisticated experiments; generating clean well-defined datasets. Reproducibility and transferability are enhanced as well as higher productivity. Furthermore, execution by non-lab scientists is enabled as well as the capacity to execute more ambitious experiments necessary to address metabolic engineering problems which would be impossible to reliably execute manually.
We present a series of case studies using Antha to exemplify what we believe is an exciting trend that could help further drive the reliability, utility and success of metabolic engineering approaches.
2. Sadowski, Michael I., Chris Grant, and Tim S. Fell. “Harnessing QbD, Programming Languages, and Automation for Reproducible Biology.” Trends in Biotechnology. doi:10.1016/j.tibtech.2015.11.006.