Publicação
Algorithms and tools for in silico design of cell factories
| Resumo: | The progressive shift from chemical to biotechnological processes is one of the pillars of the 21st century industrial biotechnology. Projections from the Organization for Economic Co-operation and Development estimate that, within the next two decades, about 35% of the production of chemicals will be guaranteed by biotechnological processes. The development of efficient cell-factories, capable of outperforming current chemical processes, is vital for this leap to happen. The development of constraint-based models of metabolism and rational computational strain optimization algorithms (CSOMs) hold the promise to accelerate these e orts. Here, we aim to provide an in depth and critical review of the currently available constraint-based CSOMs, their strengths and limitations, as well as to discuss future trends in the field. Then, we cover in detail the main tasks in strain design and provide a taxonomy of the main CSOMs. These are presented in detail and their features and limitations are explored. One of the identified problems is their limited offering of trade-o solutions of biotechnological objectives (e.g. overproducing desired compounds or minimizing the cost of solutions) versus cellular objectives (e.g. maximizing biomass). To tackle this problem we developed an evolutionary multi-objective (MO) framework for strain optimization capable of finding high-quality, trade-off solutions that can be explored by metabolic engineering experts. Also, the majority of the strain optimization algorithms rely on phenotype prediction methods based on debatable biological assumptions. We verified that, for a large percentage of solutions generated by a CSOM using one phenotype prediction method, the results would not hold when simulated with an alternative method. Leveraging on the previously developed framework and driven by the MO nature of this problem, we proposed a tandem approach capable of finding strain designs that comply with the assumptions of distinct phenotype prediction methods, validating the approach with multiple case studies. Finally, all the algorithms developed during this work are made available in the form of an open and flexible software framework. This framework is a powerful tool for both common users, interested in exploring the available methods, and experienced programmers which are able to easily extend it to support new features. |
|---|---|
| Autores principais: | Silva, Paulo Jorge Lopes Maia da |
| Assunto: | Engenharia e Tecnologia::Biotecnologia Industrial |
| Ano: | 2015 |
| País: | Portugal |
| Tipo de documento: | tese de doutoramento |
| Tipo de acesso: | acesso aberto |
| Instituição associada: | Universidade do Minho |
| Idioma: | inglês |
| Origem: | RepositóriUM - Universidade do Minho |
| Resumo: | The progressive shift from chemical to biotechnological processes is one of the pillars of the 21st century industrial biotechnology. Projections from the Organization for Economic Co-operation and Development estimate that, within the next two decades, about 35% of the production of chemicals will be guaranteed by biotechnological processes. The development of efficient cell-factories, capable of outperforming current chemical processes, is vital for this leap to happen. The development of constraint-based models of metabolism and rational computational strain optimization algorithms (CSOMs) hold the promise to accelerate these e orts. Here, we aim to provide an in depth and critical review of the currently available constraint-based CSOMs, their strengths and limitations, as well as to discuss future trends in the field. Then, we cover in detail the main tasks in strain design and provide a taxonomy of the main CSOMs. These are presented in detail and their features and limitations are explored. One of the identified problems is their limited offering of trade-o solutions of biotechnological objectives (e.g. overproducing desired compounds or minimizing the cost of solutions) versus cellular objectives (e.g. maximizing biomass). To tackle this problem we developed an evolutionary multi-objective (MO) framework for strain optimization capable of finding high-quality, trade-off solutions that can be explored by metabolic engineering experts. Also, the majority of the strain optimization algorithms rely on phenotype prediction methods based on debatable biological assumptions. We verified that, for a large percentage of solutions generated by a CSOM using one phenotype prediction method, the results would not hold when simulated with an alternative method. Leveraging on the previously developed framework and driven by the MO nature of this problem, we proposed a tandem approach capable of finding strain designs that comply with the assumptions of distinct phenotype prediction methods, validating the approach with multiple case studies. Finally, all the algorithms developed during this work are made available in the form of an open and flexible software framework. This framework is a powerful tool for both common users, interested in exploring the available methods, and experienced programmers which are able to easily extend it to support new features. |
|---|