Periodic Reporting for period 1 - BIOS (BIOS: The bio-intelligent DBTL cycle, a key enabler catalysing the industrial transformation towards sustainable biomanufacturing)
Periodo di rendicontazione: 2022-10-01 al 2024-03-31
We are all witnessing the consequences of global climate change. To prevent further rise in atmospheric CO2 levels, resources from fossil fuels should remain unused. The implementation of a circular economy is the only way to sustain a high standard of living for the current and the next generation.
To access the products of interest via bioprocesses, typically multiple heterologous genes must be implemented into microbial production hosts. Whereas synthetic biology provides a multitude of novel tools for strain engineering, their rapid and effective implementation in microbial chassis for optimal performance under industrial conditions is still very challenging.
The bio-intelligent approach in BIOS aims to accelerate and improve the conventional ‘design-build-test-learn’ (DBTL) cycle for integrated strain and bioprocess engineering underpinning. Novel innovative metrics, biosensors and bioactuators are developed and fully integrated in the automation platform. Digital twins are created mimicking cellular and process levels. By tightly intertwining AI, not only the prediction quality is improved but, crucially, hybrid learning is made possible. The power of biDBTL will be showcased by creating P. putida producer strains for terpenes, renewable-based polyesters, and methylacrylate.
To access the products of interest via bioprocesses, typically multiple heterologous genes must be implemented into microbial production hosts. Whereas synthetic biology provides a multitude of novel tools for strain engineering, their rapid and effective implementation in microbial chassis for optimal performance under industrial conditions is still very challenging.
The bio-intelligent approach in BIOS aims to accelerate and improve the conventional ‘design-build-test-learn’ (DBTL) cycle for integrated strain and bioprocess engineering underpinning. Novel innovative metrics, biosensors and bioactuators are developed and fully integrated in the automation platform. Digital twins are created mimicking cellular and process levels. By tightly intertwining AI, not only the prediction quality is improved but, crucially, hybrid learning is made possible. The power of biDBTL will be showcased by creating P. putida producer strains for terpenes, renewable-based polyesters, and methylacrylate.
The BIOS project has made significant progress across various work packages. In WP1, the project established a robust governance structure, developed a comprehensive Data Management Plan adhering to FAIR principles, and initiated the creation of the BIOS semantic communication platform. These efforts have facilitated efficient project coordination and risk management.
WP2 saw the successful start of hybrid neural-mechanistic modelling for digital twins, integrating AI and mechanistic approaches to enhance bioprocess optimization. Initial detailed kinetic models were developed, and digital twins were conceptualised for modeling transcription and translation in Pseudomonas putida, significantly advancing our understanding and control of these processes.
In WP3, advanced molecular tools, including CRISPRi, CRISPRa, and CAST, were developed and automated for high-throughput genome engineering. This enabled the precise editing and optimization of P. putida strains, leading to improved production of target compounds such as malonic acid and methacrylate.
WP4 focused on developing a plasmid-based reporter system to study gene expression regulation and automating RNA preparation and sequencing protocols, streamlining transcriptome analysis.
In WP5, adaptive evolution experiments were conducted to enhance P. putida's tolerance to project relevant stressors.. These experiments identified key stress response mechanisms and implemented strategies to improve strain robustness and productivity.
WP6 integrated machine learning algorithms with experimental data to optimise model inputs and improve strain performance. Variational autoencoders were conceptulized for efficient design of experiments, contributing to the project's hybrid learning framework.
WP7 achieved significant progress in engineering P. putida strains for the production of valuable compounds such as terpenes and tailored polymers. Promising production strains were developed, and bioprocesses were optimised using digital twins.
WP8 conducted comparative life cycle analyses to evaluate the sustainability of novel bioproduction methods. The integration of life cycle sustainability assessment into the Safe-by-Design framework ensured the project's innovations were environmentally sound.
WP9 established a comprehensive communication infrastructure to disseminate project results and engage stakeholders. Workshops and dissemination materials were created to promote the project’s achievements and foster collaboration. Following review suggestions continuing ethical qualifications were included, too.
WP2 saw the successful start of hybrid neural-mechanistic modelling for digital twins, integrating AI and mechanistic approaches to enhance bioprocess optimization. Initial detailed kinetic models were developed, and digital twins were conceptualised for modeling transcription and translation in Pseudomonas putida, significantly advancing our understanding and control of these processes.
In WP3, advanced molecular tools, including CRISPRi, CRISPRa, and CAST, were developed and automated for high-throughput genome engineering. This enabled the precise editing and optimization of P. putida strains, leading to improved production of target compounds such as malonic acid and methacrylate.
WP4 focused on developing a plasmid-based reporter system to study gene expression regulation and automating RNA preparation and sequencing protocols, streamlining transcriptome analysis.
In WP5, adaptive evolution experiments were conducted to enhance P. putida's tolerance to project relevant stressors.. These experiments identified key stress response mechanisms and implemented strategies to improve strain robustness and productivity.
WP6 integrated machine learning algorithms with experimental data to optimise model inputs and improve strain performance. Variational autoencoders were conceptulized for efficient design of experiments, contributing to the project's hybrid learning framework.
WP7 achieved significant progress in engineering P. putida strains for the production of valuable compounds such as terpenes and tailored polymers. Promising production strains were developed, and bioprocesses were optimised using digital twins.
WP8 conducted comparative life cycle analyses to evaluate the sustainability of novel bioproduction methods. The integration of life cycle sustainability assessment into the Safe-by-Design framework ensured the project's innovations were environmentally sound.
WP9 established a comprehensive communication infrastructure to disseminate project results and engage stakeholders. Workshops and dissemination materials were created to promote the project’s achievements and foster collaboration. Following review suggestions continuing ethical qualifications were included, too.
The BIOS project is pushing the boundaries of current scientific and industrial practices through several groundbreaking advancements. The integration of digital twins combining AI and mechanistic models represents a significant leap forward in predictive accuracy for strain engineering and process optimization. This innovation allows for real-time monitoring and control of bioprocesses, resulting in improved efficiency and scalability that surpass traditional methods.
Advanced genetic engineering tools, including CRISPRi, CRISPRa, and CAST, are under continuous development and automation, improving genome editing in terms of precision and throughput. These tools aim to drastically reduce the time and labour costs associated with strain development, positioning BIOS at the forefront of genetic engineering technology.
Moreover, BIOS is developing innovative bioprocesses for producing both, high-value and high-market volume compounds such as terpenes and polymers. By integrating digital twins and hybrid learning models, the project has set the framework for optimising strain engineering and bioprocessing finally to achieve high production efficiencies and to set new standards in sustainable bio-manufacturing. A key feature is the biointelligent design-build-test-learn (DBTL) cycle, as the origin of efficient strain and bioprocess engineering.
Advanced genetic engineering tools, including CRISPRi, CRISPRa, and CAST, are under continuous development and automation, improving genome editing in terms of precision and throughput. These tools aim to drastically reduce the time and labour costs associated with strain development, positioning BIOS at the forefront of genetic engineering technology.
Moreover, BIOS is developing innovative bioprocesses for producing both, high-value and high-market volume compounds such as terpenes and polymers. By integrating digital twins and hybrid learning models, the project has set the framework for optimising strain engineering and bioprocessing finally to achieve high production efficiencies and to set new standards in sustainable bio-manufacturing. A key feature is the biointelligent design-build-test-learn (DBTL) cycle, as the origin of efficient strain and bioprocess engineering.