SurreyFBA: Interactive tool for computer simulations of genome scale metabolic networks.
Lead Research Organisation:
University of Surrey
Department Name: Microbial & Cellular Sciences
Abstract
The goal of this project is to develop user-friendly, fully interactive software for computer simulation of the relationship between the genome and whole-cell metabolism.
The century of research in classical biochemistry resulted in the determination of the set of chemical conversions that enable cell to generate energy and produce its structures from nutrients available in environment. Recent revolution in molecular biology resulted in determination of the full genetic code of all organisms of interest and provided technology for monitoring genetic polymorphism and gene expression activity in experimental samples, populations and individuals. The Genome Scale Metabolic Reaction Networks (GSMNs) are computer models that integrate these resources and technologies to enable prediction of metabolic capabilities of the organism from the information about its genome and availability of nutrients in the external environment. In the process referred to as "reconstruction" the model is compiled using the set of enzymes identified by analysis of genome sequence. Each of enzymes implies presence of particular chemical reaction in the cell and the chemical reaction formula is included into the model. At the end of reconstruction process a Genome Scale Metabolic Network (GSMN) of coupled chemical reactions, linked to enzymes and genes is created. The large arsenal of so called Constrained Based Modeling (CBM) computer simulation methods is then used to explore capabilities of this network, which are consistent with chemical balances specified by the network. The CBM is an ideal approach to predict which genes affect which global metabolic capabilities, thus allowing to explore relations between genotype and phenotype of the whole cell.
Computer simulations of the relationship between genotype and whole-cell metabolism have wide existing and potential applications. In biotechnology, the CBM is already being used to rationally design genetically engineered microbial strains for production of commercially valuable substances. Existing applications include production of L-valine amino acid for cosmetics and pharmaceutical industries and production of spinosin insecticide. Increasing number of companies use designed microbial cell factories for biofuel production. The CBM is also an ideal method for detection of metabolic vulnerabilities of bacterial pathogens such as M. tuberculosis and N. meningitidis studied by the University of Surrey team. These vulnerabilities suggest enzyme inhibitors that can be used as potential new antibiotics. Finally, there is an increased research effort in application of CBM to study human metabolism. The method has huge potential to detect metabolic vulnerabilities of cancer cells and suggest drug targets. Moreover, CBM is ideal to investigate link between genetic polymorphism and metabolic syndromes, such as non-alcoholic liver disease investigated by our team.
The University of Surrey is a world leader in CBM and we have already published SurreyFBA - our own free software for computer simulation of whole cell metabolism. In this proposal we will use an 18 months long, pump priming fund to equip our software with state of the art interactive interface centered on network map editing and visualization. These features are key to enable experimental biologists to fully focus on investigation of cellular metabolism, rather than technical aspects of software usage. New generation of tools that we will champion will enable world biochemistry experts to fully use their intuition and knowledge to make fascinating discoveries leading to inventions in medicine and biotechnology.
The century of research in classical biochemistry resulted in the determination of the set of chemical conversions that enable cell to generate energy and produce its structures from nutrients available in environment. Recent revolution in molecular biology resulted in determination of the full genetic code of all organisms of interest and provided technology for monitoring genetic polymorphism and gene expression activity in experimental samples, populations and individuals. The Genome Scale Metabolic Reaction Networks (GSMNs) are computer models that integrate these resources and technologies to enable prediction of metabolic capabilities of the organism from the information about its genome and availability of nutrients in the external environment. In the process referred to as "reconstruction" the model is compiled using the set of enzymes identified by analysis of genome sequence. Each of enzymes implies presence of particular chemical reaction in the cell and the chemical reaction formula is included into the model. At the end of reconstruction process a Genome Scale Metabolic Network (GSMN) of coupled chemical reactions, linked to enzymes and genes is created. The large arsenal of so called Constrained Based Modeling (CBM) computer simulation methods is then used to explore capabilities of this network, which are consistent with chemical balances specified by the network. The CBM is an ideal approach to predict which genes affect which global metabolic capabilities, thus allowing to explore relations between genotype and phenotype of the whole cell.
Computer simulations of the relationship between genotype and whole-cell metabolism have wide existing and potential applications. In biotechnology, the CBM is already being used to rationally design genetically engineered microbial strains for production of commercially valuable substances. Existing applications include production of L-valine amino acid for cosmetics and pharmaceutical industries and production of spinosin insecticide. Increasing number of companies use designed microbial cell factories for biofuel production. The CBM is also an ideal method for detection of metabolic vulnerabilities of bacterial pathogens such as M. tuberculosis and N. meningitidis studied by the University of Surrey team. These vulnerabilities suggest enzyme inhibitors that can be used as potential new antibiotics. Finally, there is an increased research effort in application of CBM to study human metabolism. The method has huge potential to detect metabolic vulnerabilities of cancer cells and suggest drug targets. Moreover, CBM is ideal to investigate link between genetic polymorphism and metabolic syndromes, such as non-alcoholic liver disease investigated by our team.
The University of Surrey is a world leader in CBM and we have already published SurreyFBA - our own free software for computer simulation of whole cell metabolism. In this proposal we will use an 18 months long, pump priming fund to equip our software with state of the art interactive interface centered on network map editing and visualization. These features are key to enable experimental biologists to fully focus on investigation of cellular metabolism, rather than technical aspects of software usage. New generation of tools that we will champion will enable world biochemistry experts to fully use their intuition and knowledge to make fascinating discoveries leading to inventions in medicine and biotechnology.
Technical Summary
The goal of this proposal is to create a software where comprehensive set of the Constrain Based Modeling (CBM) approaches to computer simulation of Genome Scale Metabolic Reaction Networks (GSMNs) is available through fully interactive, metabolic network map centered interface.
The CBM of the GSMNs allows prediction of metabolic capabilities of the cell from repertoire of enzymes implied by genome annotation and information about media conditions. The CBM of global cell metabolism is ideal for rational design of strains in synthetic biology approach to industrial biotechnology, identification of metabolic vulnerabilities of bacterial pathogens and investigation of metabolic states of healthy, aging and diseased human tissues. The success of CBM methodology gave raise to vast array of computational protocols. Most notably, the CBM provides fundamentally new approach to analysis of omics data. Experimental data on genome scale gene expression or polymorphism are analysed in the context of whole cell metabolic model. This allows identification of global metabolic states of pathogens, cancer cells and human tissues in vivo, contributing to identification of drug targets and understanding of the genetic polymorphism in metabolic syndromes.
We have already published SurreyFBA software implementing large set of CBM methods. Here we ask for pump priming funding to equip SurreyFBA with fully interactive interface centered on metabolic network map. We will implement network editor and visualization of simulation results fully integrated with current spreadsheet-based graphics user interface. We will also improve model exchange and complement set of methods available in SurreyFBA by rFBA, OptKnock and OptCom. Thus, we will implement the most comprehensive set of CBM methods through the fundamentally new mode of user interaction, where biologists will be able to focus on using their knowledge and research intuition to generate new research hypotheses and applications.
The CBM of the GSMNs allows prediction of metabolic capabilities of the cell from repertoire of enzymes implied by genome annotation and information about media conditions. The CBM of global cell metabolism is ideal for rational design of strains in synthetic biology approach to industrial biotechnology, identification of metabolic vulnerabilities of bacterial pathogens and investigation of metabolic states of healthy, aging and diseased human tissues. The success of CBM methodology gave raise to vast array of computational protocols. Most notably, the CBM provides fundamentally new approach to analysis of omics data. Experimental data on genome scale gene expression or polymorphism are analysed in the context of whole cell metabolic model. This allows identification of global metabolic states of pathogens, cancer cells and human tissues in vivo, contributing to identification of drug targets and understanding of the genetic polymorphism in metabolic syndromes.
We have already published SurreyFBA software implementing large set of CBM methods. Here we ask for pump priming funding to equip SurreyFBA with fully interactive interface centered on metabolic network map. We will implement network editor and visualization of simulation results fully integrated with current spreadsheet-based graphics user interface. We will also improve model exchange and complement set of methods available in SurreyFBA by rFBA, OptKnock and OptCom. Thus, we will implement the most comprehensive set of CBM methods through the fundamentally new mode of user interaction, where biologists will be able to focus on using their knowledge and research intuition to generate new research hypotheses and applications.
Planned Impact
Computational tools developed in this project will benefit two major groups of users in two major application areas of predictive biology. First, the Industrial Biotechnology (IB) users will be immediately able to apply the software developed in this project for design of strains, cell lines and bioprocesses. Second, in longer timescale, the tools will find essential applications in the area of Predictive, Preventive and Personalised (3P) Medicine making an impact in health, healthy aging and quality of life. Eventually, our computational tools will be used in Pharmaceutical, Food and Chemical industries.
The fully interactive, metabolic map centered interface to comprehensive set of computational protocols, delivered by this project, will attract new users who will benefit from predictive simulations in Constrained Based Modelling (CBM) framework. The CBM enables predictive simulation of Genome Scale Metabolic Networks (GSMNs), which has already made economic impact in Industrial Biotechnology. For example Genomatica (www.genomatica.com) offers sustainable chemical manufacturing technologies designed by CBM of GSMNs. Another success story is a complete synthethic biology pipeline in Dow Agro (M. Donahue, ICSB 2010 presentation) involving microbial genome sequencing, GSMN reconstruction and CBM leading to rationally designed strains with increased spinosin production. However, the tools used by these companies require substantial investment into expert staff dedicated to operation of technically complex computational tools. Our desktop software will bring full power of CBM to experimental biologists thus reducing the costs of using this computational approach. This will especially benefit SMEs and spin-off companies with limited staff numbers. In these companies experimental scientists involved in innovative research will be able to use predictive computer simulations to support design of new strains.
Complex multifactorial diseases have enormous impact on economy, quality of life and healthy aging. For example, disruption of metabolic networks controlling nutrient metabolism and energy homeostasis in the liver leads to conditions such as obesity, fatty liver and type 2 diabetes. In 2007 24% of adults and 17% of children in the UK were obese and the estimated costs of treating the consequences of obesity were £1 billion in 2002 and projected to be £5.3 billion by 2025 (NHS Information Centre, Lifestyle Statistics, 2009). These tremendous social and economic costs can only be decreased if disease can be prevented before it has to be cured. Scientific community is now in the position to develop tools for using patient's genetic information to provide personal advice on nutrition and lifestyle leading to disease prevention. Mechanistic understanding of genome scale metabolic networks is necessary to realise the 3P Medicine vision. Current CBM tools are already yielding important predictions of the genetic polymorphism effects in healthy and diseased tissue metabolism. Our biologist-oriented implementation will accelerate adoption of large-scale network modeling by both industry and academia.
Finally, University of Surrey will reinforce its position of the center of excellence in GSMN modeling and will offer unique training to PDRAs, postgraduate and graduate students providing highly trained staff to the UK economy. Early adoption of state-of-the art predictive biology approaches will contribute to the international competitiveness of UK Pharmaceutical, Food and Chemical industries. We will actively seek to demonstrate new SurreyFBA tools to the existing industry contacts of Co-Is in GSK, Pfizer, AstraZeneca (NJP), Sanofi-Pasteur (JMcF) and industrial partners of BBSRC's IBTI Club (CAR).
The fully interactive, metabolic map centered interface to comprehensive set of computational protocols, delivered by this project, will attract new users who will benefit from predictive simulations in Constrained Based Modelling (CBM) framework. The CBM enables predictive simulation of Genome Scale Metabolic Networks (GSMNs), which has already made economic impact in Industrial Biotechnology. For example Genomatica (www.genomatica.com) offers sustainable chemical manufacturing technologies designed by CBM of GSMNs. Another success story is a complete synthethic biology pipeline in Dow Agro (M. Donahue, ICSB 2010 presentation) involving microbial genome sequencing, GSMN reconstruction and CBM leading to rationally designed strains with increased spinosin production. However, the tools used by these companies require substantial investment into expert staff dedicated to operation of technically complex computational tools. Our desktop software will bring full power of CBM to experimental biologists thus reducing the costs of using this computational approach. This will especially benefit SMEs and spin-off companies with limited staff numbers. In these companies experimental scientists involved in innovative research will be able to use predictive computer simulations to support design of new strains.
Complex multifactorial diseases have enormous impact on economy, quality of life and healthy aging. For example, disruption of metabolic networks controlling nutrient metabolism and energy homeostasis in the liver leads to conditions such as obesity, fatty liver and type 2 diabetes. In 2007 24% of adults and 17% of children in the UK were obese and the estimated costs of treating the consequences of obesity were £1 billion in 2002 and projected to be £5.3 billion by 2025 (NHS Information Centre, Lifestyle Statistics, 2009). These tremendous social and economic costs can only be decreased if disease can be prevented before it has to be cured. Scientific community is now in the position to develop tools for using patient's genetic information to provide personal advice on nutrition and lifestyle leading to disease prevention. Mechanistic understanding of genome scale metabolic networks is necessary to realise the 3P Medicine vision. Current CBM tools are already yielding important predictions of the genetic polymorphism effects in healthy and diseased tissue metabolism. Our biologist-oriented implementation will accelerate adoption of large-scale network modeling by both industry and academia.
Finally, University of Surrey will reinforce its position of the center of excellence in GSMN modeling and will offer unique training to PDRAs, postgraduate and graduate students providing highly trained staff to the UK economy. Early adoption of state-of-the art predictive biology approaches will contribute to the international competitiveness of UK Pharmaceutical, Food and Chemical industries. We will actively seek to demonstrate new SurreyFBA tools to the existing industry contacts of Co-Is in GSK, Pfizer, AstraZeneca (NJP), Sanofi-Pasteur (JMcF) and industrial partners of BBSRC's IBTI Club (CAR).
Organisations
Publications
Thumser AE
(2014)
Fatty acid binding proteins: tissue-specific functions in health and disease.
in Current opinion in clinical nutrition and metabolic care
Wu H
(2016)
MUFINS: multi-formalism interaction network simulator
in npj: Systems Biology and Applications
Wu H
(2016)
MUFINS: multi-formalism interaction network simulator.
in NPJ systems biology and applications
| Description | Constraint-based modelling (CBM) enables mechanistic simulation of genotype-phenotype relationship in whole-cell metabolism. It is one of the major modelling approaches in Synthetic Biology, where it is used for rational design of microbial strains producing commercially valuable chemicals. There is a growing interest in using CBM to study how cellular regulatory processes reprogram metabolism, especially in the context of tissue specific models in health and disease. We developed SurreyFBA2, which provides graphics user interface (GUI) to both classical CBM and simulation of metabolic reprogramming by regulatory processes. We implement top-down approaches of deriving metabolic states from ~omics data, including our new variant of congruency approach. The software also provides bottom-up methods where regulation is mechanistically modelled. The linear inhibition constraint enables for the first time linear modelling of networks simultaneously describing gene regulation, signalling and metabolism at steady state. An extended version of our Quasi Steady State Petri Net approach allows integration of dynamic models with CBM. This allows user to integrate dynamic models of gene regulatory circuits and cell signalling with CBM of whole-cell metabolism. Microbial community models, where different species compete for nutrients as well as integrated models of bioreactor flows and cellular metabolism can be simulated. The GUI supports editing of metabolic, regulatory and integrated models, display of results and visualisation of metabolic fluxes on network maps with automatic and editable layouts. Stand-alone command line tool is available for integration with bioinformatics pipelines. We have also developed first version of Simulocyte, a web interface to SurreyFBA, which visualises genome scale molecular networks in interactive 3D interface. Simulocyte will allow formation of online communities developing molecular network models. As a part of user engagement activities, the prototype of Simulocyte interface has been presented to C1Net Network in Industrial Biotechnology and Bioenergy (NIBB). We have received pump priming funding to initiate community-based modelling of Methanotroph metabolism. We expect the first version of the model to be available for C1Net community in September 2015. |
| Exploitation Route | The software developed in this project will enable community based reconstruction of genome scale molecular interaction networks. These models will be used for prediction of genotype phenotype relationship in a variety of fields including stratified medicine and synthetic biology. In synthetic biology area we have already engaged with C1Net Network for Industrial Biotechnology and Bioenergy (http://www.c1net.co.uk). We have received pump priming fund to apply Simulocyte for development of predictive model used for design of new bacterial strains converting industrial waste gases to commercially valuable chemicals. We are also exploiting opportunities of applying our software in the context of toxicity predictions for pharmaceutical industry (Surrey TTO office contacts). |
| Sectors | Agriculture, Food and Drink,Chemicals,Healthcare,Pharmaceuticals and Medical Biotechnology |
| URL | http://sysbio3.fhms.surrey.ac.uk/sfba/index.html |
| Description | The SurreyFBA2 software developed in this project will enable simulation of genome scale molecular interaction networks. These models will be used for prediction of genotype phenotype relationship in a variety of fields including stratified medicine and synthetic biology. We expect to achieve economic impact through collaboration with C1Net NIBB network, where we won pump priming fund for development of modelling platform enabling synthetic biology approach. Through Surrey TTO office we exploit opportunities of creating a spin-off company providing in-vivo/in-vitro extrapolation service and bespoke software application for pharmaceutical industry. |
| First Year Of Impact | 2014 |
| Sector | Agriculture, Food and Drink,Chemicals,Environment,Pharmaceuticals and Medical Biotechnology |
| Impact Types | Economic |
| Title | SurreyFBA2: Software for multi-formalism simulation of genome scale molecular interaction networks. |
| Description | We developed SurreyFBA2, which provides graphics user interface (GUI) to both classical CBM and simulation of metabolic reprogramming by regulatory processes. We implement top-down approaches of deriving metabolic states from ~omics data, including our new variant of congruency approach. The software also provides bottom-up methods where regulation is mechanistically modelled. The linear inhibition constraint enables for the first time linear modelling of networks simultaneously describing gene regulation, signalling and metabolism at steady state. An extended version of our Quasi Steady State Petri Net approach allows integration of dynamic models with CBM. This allows user to integrate dynamic models of gene regulatory circuits and cell signalling with CBM of whole-cell metabolism. Microbial community models, where different species compete for nutrients as well as integrated models of bioreactor flows and cellular metabolism can be simulated. The GUI supports editing of metabolic, regulatory and integrated models, display of results and visualisation of metabolic fluxes on network maps with automatic and editable layouts. Stand-alone command line tool is available for integration with bioinformatics pipelines. |
| Type Of Technology | Software |
| Year Produced | 2015 |
| Open Source License? | Yes |
| Impact | SurreyFBA2 stems from legacy of research on simulation of genome scale molecular interaction networks at the University of Surrey. Previous versions of the software have been used to reconstruct genome scale metabolic networks of bacterial pathogens Mycobacterium tuberculosis and Neisseria meningitidis. In biotechnology context, we applied SurreyFBA to simulations of Streptomyces coelicolor - major industrial microorganism for antibiotic production. Currently, we use SurreyFBA to study metabolic reprogramming in non-alcoholic liver disease and cancer. The work on cancer cell metabolism attracted industry contribution through BBSRC CASE PhD fellowships. External users applied SurreyFBA as a computational engine in their metabolomics website and as a research tool. We have also attracted interest of industrial and academic partners of C1Net Network of Industrial Biotechnology and Energy. The SurreyFBA2 is used to develop model of Methanotroph. |
| URL | http://sysbio3.fhms.surrey.ac.uk/sfba/index.html |