Centre for Systems Biology at Edinburgh

Lead Research Organisation: University of Edinburgh
Department Name: Sch of Biological Sciences


Systems Biology is a fascinating development in modern biology. We have achieved a good general understanding of how cells work, including the 'central dogma': genes are transcribed to RNA; a splicing process produces mature messenger RNA; mRNA is translated to proteins; and protein pathways regulate gene expression and perform other functions such as detecting intercellular signals. With the Human Genome Project we know our DNA sequence and have a partial map of our genes. And, finally, high-throughput biology is giving us massive amounts of time series data, e.g., of protein or mRNA concentrations. Systems Biology seeks to understand how biological systems function by integrating all this knowledge. System theories are implemented as in silico models: computer simulations built using mathematical models. Biological systems are extraordinarily complex with many levels of interacting subsystems. We therefore expect to construct models by combining submodels, beginning with pathways, and eventually proceeding to organelles, cells, physiological systems and whole organisms. One hopes to be able to predict the effect of variations, e.g.: environmental, or resulting from disease or adding drugs. Such a Systems Biology would produce an enormous increase in understanding and lead to major progress in medicine, agriculture and industry. The Edinburgh Centre for Systems Biology will advance our ability to make and use such models by applying advanced computer science and mathematical techniques to a carefully chosen range of important biological systems which are different enough to test our model-making ability to the limit. Our largest such system is the interferon pathway, an important signaling pathway in macrophages, the main immune system cells. It is already hard here to conveniently describe the pathway intricacies, and we shall develop new international graphical standards. The middle-sized system is RNA metabolism, the process leading from raw to mature RNA. It should be possible to model this still complex system in detail, correlating the models with high-throughput data. Finally comes circadian rhythm, biological clocks, where small genetic circuits regulate large parts of gene expression. Here we may perform mathematical analyses, e.g., investigating how light and temperature synchronise clocks in a noisy environment. The traditional modelling technique of mathematical biology uses systems of differential equations: systems biology presents new challenges. We shall produce SBSI, a modelling facility of industrial quality freely available to all. Probabilistic models are sometimes more realistic than differential ones, e.g., for few protein molecules. We shall explore such variations to ensure realistic yet tractable modelling. High-throughput data are noisy and hard to obtain in sufficient quantity. We shall apply Bayesian techniques, familiar from Artificial Intelligence, to help discover pathways. We wish to construct big systems from small ones (modules) and to experiment efficiently with system variants. Programming languages let one do this for computational systems; we shall apply the lessons learnt to design and use languages for biological ones, with the additional prospect of being able to query and design systems using special logics. In summary, we intend to build a science of Systems Biology using computer science and mathematics to produce models refined by and informing biological experiment. The variety of the biology we do will ensure the wide usefulness of the techniques; the variety of the techniques we will explore will give the enterprise every prospect of success. However to achieve usefulness requires much more. We will therefore combine our scientific effort with training and outreach programmes: the one to contribute to the production of the next generation of systems biologists; and the other to make our work available to our colleagues in academia and our partners in industry.

Technical Summary

The Centre for Systems Biology will create a unique multidisciplinary environment, as part of the University's vision 'to integrate the physical with the life sciences'. The University has targeted new infrastructure and staff into interdisciplinary biology and is committed to a new building for the Centre. The Centre's research will focus on modelling dynamic biological systems, rather than concentrating on a single species, organ, or biological process. Thus we face the intellectual challenge of biological modelling in its entirety, not through the lens of an individual biological question. We will develop world-leading modelling methods based on our current work and informed by the concrete requirements of our pilot biological projects. The pilot projects are selected to cover a range of dynamic complexity, timescales, component numbers and current modelling: they focus on RNA metabolism, interferon signalling, and circadian rhythms. We will advance biological understanding in each of these areas by combining experimental and modelling approaches, including a high-quality experimental core facility. Comparing across species and biological processes, we will abstract broad design principles of biological structure and dynamics. The Centre will benefit from our current centres of excellence, especially in functional genomics (with HTP facilities) and high-performance computing (with IBM BlueGene supercomputer). However, we cannot cover all aspects of modelling with CISB funding. We will use this seed funding to establish infrastructures that are broadly applicable to any systems biology project. The automated software infrastructure in particular, a world first in academia, will allow our international collaborators and the wider community to add applications in a modular fashion. Future funding, some at advanced stages of negotiation, will extend the Centre's research, consolidating its position as an international centre of excellence in systems biology.


10 25 50
Description SynthSys research directly funded by this award published 110 primary papers and 13 reviews, generating many new theoretical tools and experimental methods, as well as revealing for the first time how yeast transcription pauses at a splicing checkpoint, how mouse macrophage cells downregulate cholesterol synthesis as an antiviral defense, and that eukaryotic cells have a paradigm-shifting, non-transcriptional 24-hour clock.

SynthSys created durable research platforms for biochemistry and quantitative proteomics (the Kinetic Parameter Facility, KPF), for informatics (SBSI and online repositories BioDare and PlaSMo), and for interdisciplinary research management.

SynthSys created a new, sustainable Centre for interdisciplinary biology, attracting £22.4M further external funding and new academic staff, and thus doubling the number of Systems Biology research groups at the University of Edinburgh.
Exploitation Route Many collaborators and new projects have already arisen through creation of the Centre, in addition to new projects and results arising from the work in the original groups.
Sectors Aerospace, Defence and Marine,Agriculture, Food and Drink,Digital/Communication/Information Technologies (including Software),Education,Energy,Healthcare,Manufacturing, including Industrial Biotechology,Pharmaceuticals and Medical Biotechnology
URL http://www.synthsys.ed.ac.uk
Description The impacts from this large award were described in full in the final report. A major recent funding award to Edinburgh for a Research Centre in Synthetic Biology was an additional impact.
Description SynthSys: policy impacts
Geographic Reach National 
Policy Influence Type Participation in a national consultation
Impact Largely through Directorship of CSBE/Assoc. Directorship of SynthSys, Millar has participated in national and international working groups and policy forums:BBSRC Digital organisms strategy working group, 2012
Title Mathematical model of Arabidopsis vegetative growth: the Framework Model (v1; Chew et al PNAS 2014) 
Description Mathematical model of Arabidopsis vegetative growth: the Framework Model General information State: Published Organisations: School of Biological Sciences Authors: Chew, Y. H., Millar, A. Publication date: 2014 Publication information Media of output: online file Year: 2014 Original language: English Links: http://www.plasmo.ed.ac.uk/plasmo/models/model.shtml?accession=PLM_76 Research output: Other contribution PublishedFor validation 
Type Of Material Computer model/algorithm 
Year Produced 2014 
Provided To Others? Yes  
Impact Mathematical model of Arabidopsis vegetative growth: the Framework Model General information State: Published Organisations: School of Biological Sciences Authors: Chew, Y. H., Millar, A. Publication date: 2014 Publication information Media of output: online file Year: 2014 Original language: English Links: http://www.plasmo.ed.ac.uk/plasmo/models/model.shtml?accession=PLM_76 Research output: Other contribution PublishedFor validation 
URL http://www.plasmo.ed.ac.uk/plasmo/models/model.shtml?accession=PLM_76
Title Mathematical model of the Arabidopsis circadian clock: Pokhilko 2012 model 
Description The model is termed P2012 and derives from the article: Modelling the widespread effects of TOC1 signalling on the plant circadian clock and its outputs. Alexandra Pokhilko, Paloma Mas & Andrew J Millar BMC Syst. Biol. 2013; 7: 23, submitted 10 Oct 2012 and published 19 March 2013. The model describes the circuit depicted in Fig. 1 of the paper. It updates the P2011 model from Pokhilko et al. Mol. Syst. Biol. 2012 model, PLM_64, by including: TOC1 as a repressor of multiple clock genes, rather than only of LHY/CCA1. ABAR transcription modified by TOC1, affecting stomatal aperture. TOC1 transcription modified by ABA. SBML curation notes on PlaSMo (please see Comments for each version): PLM_70 version 1 is the version published as Supplementary Information and submitted to the Biomodels database. Copasi and MATLAB versions are attached to version 1. General information State: Published Organisations: School of Biological Sciences Authors: Pokhilko, A., Millar, A. Publication date: 2013 Publication information Media of output: Database records, various formats Year: 2013 Original language: English Links: http://www.plasmo.ed.ac.uk/plasmo/models/model.shtml?accession=PLM_70 http://www.ebi.ac.uk/compneur-srv/biomodels-main/BIOMD0000000445 
Type Of Material Computer model/algorithm 
Year Produced 2012 
Provided To Others? Yes  
Impact This model has formed the basis for further work by external groups, as described in the following links: Fogelmark et al. PLoS CB 2014 - http://dx.doi.org/10.1371/journal.pcbi.1003705 Zhou et al. Nature 2015 - http://dx.doi.org/10.1038/nature14449 Foo et al. PLoS CB 2016 - http://dx.doi.org/10.1371/journal.pcbi.1004748 Calluwe et al. Frontiers 2016 - http://dx.doi.org/10.3389/fpls.2016.00074 
URL http://www.plasmo.ed.ac.uk/plasmo/models/model.shtml?accession=PLM_70
Title Mathematical model of the Arabidopsis circadian clock: the P2011 model 
Description [from the record in PlaSMo] This model is termed P2011 and derives from the article: The clock gene circuit in Arabidopsis includes a repressilator with additional feedback loops. Alexandra Pokhilko, Aurora Piñas Fernández, Kieron D Edwards, Megan M Southern, Karen J Halliday & Andrew J Millar Mol. Syst. Biol. 2012; 8: 574, published 6 March 2012. Link The model describes the circuit depicted in Fig. 1 of the paper (GIF attached). It updates the Pokhilko et al. 2010 model (termed P2010), PLM_6, by including: the Evening Complex genes (ELF4, ELF3, LUX), light-regulated degradation of ELF3 by COP1, TOC1 as a repressor rather than an activator of LHY/CCA1.These changes allowed the removal of hypothetical components TOC1mod (or X) and Y from the earlier models. They also reveal that the central loop of the model is a triple-repressor ring oscillator, or 'repressilator' (illustrated in Fig. 8, GIF attached). Compared to the model version submitted to the Biomodels database, this version slightly alters the names of some variables and uses an SBML AssignmentRule for the light input. This will facilitate the use of a generic SBML StepFunction to better describe the light-dark cycle, in a subsequent version of this model. 
Type Of Material Computer model/algorithm 
Year Produced 2012 
Provided To Others? Yes  
Impact Further publications and model development 
Title Mathematical model of the clock gene circuit in the alga Ostreococcus tauri 
Description Circadian clocks are biological timekeepers that allow living cells to time their activity in anticipation of predictable environmental changes. Detailed understanding of the circadian network of higher plants, such as Arabidopsis thaliana, is hampered by the high number of partially redundant genes. However, the picoeukaryotic alga Ostreococcus tauri, which was recently shown to possess a small number of non-redundant clock genes, presents an attractive alternative target for detailed modelling of circadian clocks in the green lineage. Based on extensive time-series data from in vivo reporter gene assays, we developed a model of the Ostreococcus clock as a feedback loop between the genes TOC1 and CCA1. The model reproduces the dynamics of the transcriptional and translational reporters over a range of photoperiods. Surprisingly, the model is also able to predict the transient behaviour of the clock when the light conditions are altered. Despite the apparent simplicity of the clock circuit, it displays considerable complexity in its response to changing light conditions. Systematic screening of the effects of altered day length revealed a complex relationship between phase and photoperiod, which is also captured by the model. The complex light response is shown to stem from circadian gating of light-dependent mechanisms. This study provides insights into the contributions of light inputs to the Ostreococcus clock. The model suggests that a high number of light-dependent reactions are important for flexible timing in a circadian clock with only one feedback loop. 
Type Of Material Computer model/algorithm 
Year Produced 2011 
Provided To Others? Yes  
Impact Research funding and publications 
URL http://www.plasmo.ed.ac.uk
Title Model for starch polymerisation and degradation in the alga, Ostreococcus tauri. 
Description BACKGROUND: The storage of photosynthetic carbohydrate products such as starch is subject to complex regulation, effected at both transcriptional and post-translational levels. The relevant genes in plants show pronounced daily regulation. Their temporal RNA expression profiles, however, do not predict the dynamics of metabolite levels, due to the divergence of enzyme activity from the RNA profiles.Unicellular phytoplankton retains the complexity of plant carbohydrate metabolism, and recent transcriptomic profiling suggests a major input of transcriptional regulation. RESULTS: We used a quasi-steady-state, constraint-based modelling approach to infer the dynamics of starch content during the 12 h light/12 h dark cycle in the model alga Ostreococcus tauri. Measured RNA expression datasets from microarray analysis were integrated with a detailed stoichiometric reconstruction of starch metabolism in O. tauri in order to predict the optimal flux distribution and the dynamics of the starch content in the light/dark cycle. The predicted starch profile was validated by experimental data over the 24 h cycle. The main genetic regulatory targets within the pathway were predicted by in silico analysis. CONCLUSIONS: A single-reaction description of starch production is not able to account for the observed variability of diurnal activity profiles of starch-related enzymes. We developed a detailed reaction model of starch metabolism, which, to our knowledge, is the first attempt to describe this polysaccharide polymerization while preserving the mass balance relationships. Our model and method demonstrate the utility of a quasi-steady-state approach for inferring dynamic metabolic information in O. tauri directly from time-series gene expression data. 
Type Of Material Computer model/algorithm 
Year Produced 2011 
Provided To Others? Yes  
Impact Further publications, model development in Arabidopsis. 
Title The Biodare Data Repository 
Description BioDare, an integrated data analysis and sharing resource for dynamic biological systems, by Zielinski, Moore, Troup, Beaton, Adams, Halliday, Millar. 2016 overview: BioDare returns immediate value to any user who uploads data, directly justifying the time that they spend in describing and organising their data. This makes BioDare unusual among biological data management systems. It is entirely typical that this immediate value is highly targeted, to users who require specialised analysis of rhythmic data. In addition, it facilitates data sharing and public dissemination, which give value in the much longer term. 2011 summary: BioDare, was developed to store, share and analyse rhythmic time series data. Currently it stores more than 70000 time series with over 9 million time points. The repository supports the description and processing of data from various experimental techniques, as well as literature data. It allows searching and aggregation of data from independent experiments and subsequent visualisation of not only original data but also processed data (averaged, normalized, detrended). BioDare also performs data analysis by executing period analysis routines via web services, including FFT-NLLS, mFourfit and the ROBuST spectrum resampling method. BioDare was designed initially to support the ROBuST project [opening to ROBuST users in 2009], and was extended for SynthSys and TiMet projects. It is highly relevant to other similar research, worldwide. The data infrastructure team is following a staged process to open the data repository and associated web services for analysis of rhythmic data to external users. Six potential beta-testing locations were recruited and visited in Jan-Feb 2011. Requirements specified by these betatesters have been progressively included in the system, in some cases over multiple rounds of interaction. Further beta-test users were recruited in the summer of 2011. We expect to open the system to additional users in the Spring of 2012, and to make the system public within the year. 2016 update. BioDare was made public as proposed and additional external users were recruited at scientific conferences in 2012-2014, including the UK circadian clock clubs, Gordon conferences on Chronobiology and GARNET data management workshops. BioDare's data analysis was transformed to support public use. First SynthSys, then in 2015 the UK Centre for Mammalian Synthetic Biology provided upgraded computer servers. Both the original analysis methods and four further rhythm analysis methods were refactored to native Java, greatly enhancing compute speed and stability, in part through a collaborative project with Edinburgh's supercomputing centre EPCC (see Zielinski et al. 2014 for detailed method evaluation and user guidance). The detailed experimental metadata required from users now supports a very powerful search method, which aggregates data from multiple labs and experiments. Data visualisation is more flexible, with many secondary data series (normalised, de-trended, averages, error bars, etc) pre-computed for rapid graphical display. Any data displayed can be downloaded as a numerical spreadsheet, to reproduce exactly the online graphs. As of February 2015, BioDare held over 41 million data points, in 232,844 timeseries, from 2,344 experiments. The 10 largest user labs were from UK, USA, Chile and Sweden. The largest single user lab by experiments works on circadian clocks in mouse cell and tissue cultures, at MRC LMB, Cambridge UK. The largest user lab by timeseries is from the original ROBuST project, working on plant circadian clocks. (see Flis et al. 2015). Partial cost recovery started in 2014: heavy users of data analysis functions pay an annual subscription. To encourage data sharing, users who release their BioDare data for public dissemination gain "analysis credits", which can fully support their usage costs. 
Type Of Material Database/Collection of data 
Year Produced 2009 
Provided To Others? Yes  
Impact Please see the activities and publications linked to the relevant awards in ResearchFish: - workshops as noted in description above. - publications describing aspects of BioDare: Zielinski et al. 2014; Moore et al. 2014; Flis et al. 2015. - publications using BioDare for data dissemination include Gould et al. 2013; Flis et al. 2015; Millar et al. 2015. There is no point in re-entering all the data here, that's the point of ResearchFish. 
URL http://www.biodare.ed.ac.uk
Title The Plant Systems Modelling portal, PlaSMo 
Description The Plant Systems Modelling portal, PlaSMo, meets a need identified as a priority by the plant systems biology community during a workshop on Succeeding in Plant Systems Biology organised by BBSRC and GARNet, the Genomic Arabidopsis Resource Network, in July 2005. Existing types of models used by systems biologists, which are suitable for representing metabolic pathways and genetic regulatory networks at the subcellular and cellular level, lack the properties necessary for modelling the growth and environmental responses of whole plants. To address this concern, a further meeting bringing together researchers from the systems biology, bioinformatics and agro-ecological modelling communities (Interfacing Systems Biology with Crop and Ecosystem Modelling; Swindon, March 2006) was jointly organised by GARNet and crop scientists from IGER (now IBERS) Aberystwyth. There was a strong consensus at the meeting that existing crop models, some produced as long ago as the 1960s and 70s, together with more recent approaches such as L-systems modelling of plant morphogenesis, have enormous potential to contribute to plant systems biology. However, it was acknowledged that much expertise in crop modelling has suffered attrition in the last two decades, and that many of the models developed by these experts were in danger of being lost A key recommendation in the final report of the workshop was therefore the development of: "a Plant Specific Model Repository - a central resource of current plant based models for others to use and test." PlaSMo is the response. 
Type Of Material Database/Collection of data 
Year Produced 2010 
Provided To Others? Yes  
Impact Multiple subsequent projects and publications have used this resource. 
URL http://www.plasmo.ed.ac.uk