samedan logo
home > ebr > summer 2013 > all systems go
European Biopharmaceutical Review

All Systems Go

In recent years, systems biology (SB) has arisen as a new discipline that aims to understand cell complexity by considering the available knowledge as a global system where the specific elements are less important than the global relationship between them (1). The maps generated by SB represent nature using two main strategies: those based on the topological information of the map (topological); and those that include the map but also any other functional information associated to a cell (functional).

Using these approaches, it is possible to identify crucial points over the map that eventually become key proteins in cells, clarify the mechanisms of action and conduct drug repositioning (2-6). Unfortunately, the complexity of cells and the low quality of available data are two huge impediments to carrying out successful analyses. It is then necessary to create simpler models to determine the way of analysing the graph that represents the cell.

Virtual Worm Study

To that end, the following article represents a virtual toy model of a worm with basic vital functions of growth, feeding, excretion, respiration and temperature regulation. A combination of topological and functional strategies has been applied in order to understand how the virtual worm lives. By deepening our insight on a toy model’s behaviour, we take another step towards understanding real life. Ultimately, this added level of comprehension may lead to new and exciting findings, such as improved drug design, personalised treatments or adverse events prediction. To meet the objectives planned for this work, a model of a cell was defined and then a mathematical strategy was applied to understand its behaviour (7).

Creating a Virtual Life

The cell model consists of a graph where four pathways associated to basic functions were defined as sets of genes or proteins. Figure 1 describes the four pathways from a topologic point of view. Two types of relevant genes/proteins in the map were selected by the virtual worm designers on the basis of current knowledge before starting any analysis. Seed proteins are indispensable proteins that have a relevant role in any of the four pathways: P3, P5, P11, P23 and P28. Effector proteins are nodes with measurable function associated to a pathway activity: P8, P14, P24 and P30.

The functional principles of life of the virtual worm have been described as a truth-table that contains definitions about the four pathways and how they interact with the environment. In order to represent abstract biological information in the map, four additional nodes were added as the abstraction of the function of each pathway.

Understanding the Virtual Organism Life Cycle

Once the model is created, it is necessary to define a mathematical strategy to understand it. Two different strategies were applied. Table 1 contains information about their requirements and inputs.

● The topological analysis studies the structure of genes or proteins of the virtual worm: the virtual worm itself. By discovering highly connected proteins (hubs) and nodes with the highest number of pathways going through them (bottlenecks), it allows the identification of novel effector proteins

● The functional analysis studies the relation of the virtual worm with the environment and how the virtual worm adapts its genes/ proteins to its life conditions. The objective is to create a mathematical algorithm that complies with all the truth-table information with the restriction given by the topology of genes/proteins

Table 1 shows the available and hidden information, alongside the results and validation of the topological and functional analyses.

Results of the Analyses

To understand how the virtual worm lives is equivalent to discovering what the key genes or proteins are and how the virtual worm interacts with its environment.

The objective of the topological analysis is to identify effector proteins in the gene or protein map. As shown in Figure 2a (page 37), the effector proteins obtained were P8, P14, P24 and P30 – the same as previously defined by the virtual worm designer.

It is necessary to perform the functional analysis in order to determine the seed proteins. Following this procedure, genes or proteins P3, P11, and P23 were obtained as relevant genes or proteins for the flux of information over the map, coinciding with those selected by the worm designers (see Figure 2b, page 37).

Additionally, functional analysis is employed to predict effector proteins. As shown in Figures 3a, 3b, 3c and 3d, in three of the four pathways functional analysis has reached the same results as the topological analysis and the effectors defined by the virtual worm designer. The only exception is the temperature control pathway (see Figure 3d).

In a deeper analysis of the mathematical algorithm, it is possible to extract more detailed conclusions about the virtual worm’s life. For instance, the presence of food and oxygen activates the pathways of respiration and feed: P11 is being produced, being an essential role to synthesise P20, P21 and P22 metabolites of the excretion pathway.


This project recreates the life of a virtual worm defined exclusively by the relation between its genes/ proteins and its way of living. Two different strategies were applied to understand the life of the virtual worm and determine any knowledge previously unsuspected by its designers – topological and functional.

By comparison of the results, we can confirm that both methods came to similar conclusions: 75 per cent of effector proteins determined by the topological analysis have been confirmed as effector proteins in the functional analysis (see Figures 3a, 3b, 3c and 3d). Moreover, the seed proteins provided by the virtual worm designers as requirement for the topological analysis were also confirmed by the functional analysis (see Figure 3b). Thus, it is possible to comply with the requirements of the topological analysis using the list of seed proteins obtained from the functional analysis. Therefore, it is not necessary to have a broader knowledge about a species to reach some conclusions about its life.

Future Perspective and Applications

By means of the functional analysis, it is possible to obtain some answers from the mathematical algorithm that contains a summary of the virtual worm’s life. For example, it would be possible to design a personalised treatment or to predict adverse events of a drug.

Unfortunately, this strategy is not yet applicable to humans, where the number of nodes and links is too high for current calculation capacity. A possible solution to this limitation is to simplify the information by aggregating the proteins or increasing the abstraction of the nodes (8). Alternatively, simple organisms, such as bacteria, pose as good candidates for the application of these techniques. In this manner, mechanisms of infection could be identified or new antibiotics could be designed.

There is an enormous interest in this area and some disciplines have been trying to model real life (9-12). In a purely biologic context, the best modelling lies in an integrative approach: gene regulation, metabolism, environmental stimulation or response, and their interconnections have to be included (13).


This article was produced with the help of Raquel Valls, Vesna Radovanovic, Emre Guney, Javier Garcia-Garcia, Victor Codony Domenech, Laura Corredor Gonzàlez and Baldo Oliva The research leading to these results was co-funded by the European Union’s Seventh Framework Programme (FP7/2007-2013) under grant agreements no HEALTH-F3-2009-223101 (AntiPathoGN) and no HEALTH-F2-2010-261460 (GUMS&JOINTS). We also thank our partners of the SHiPrecERASysBio+ initiative (EUI-2009- 04018).


1. Kitano H, Systems biology: a brief overview, Science 295(5560): pp1,662-1,664, 2002

2. Hase T et al, Structure of protein interaction networks and their implications on drug design, PLoS Comput Biol 5(10): ppe1000550, 2009

3. Sambourg L and Thierry-Mieg N, New insights into protein-protein interaction data lead to increased estimates of the S. cerevisiae interactome size, BMC Bioinformatics 11: pp605, 2010

4. Yoshida M et al, Diagnosis of gastroenterological diseases by metabolome analysis using gas chromatography-mass spectrometry, J Gastroenterol, 47(1): pp9-20, 2012

5. Ashburn TT and Thor KB, Drug repositioning: identifying and developing new uses for existing drugs, Nat Rev Drug Discov, 3(8): pp673-683, 2004

6. Dudley JT, Deshpande T and Butte AJ, Exploiting drug-disease relationships for computational drug repositiong, Brief Bioinform, 12(4): pp303-311, 2011

7. Pujol A, BO et al, Virtual-organism toymodel as a tool to develop bioinformatics approaches of Systems Biology for medical-target discovery. Visit:

8. Mas JM, PA, Aloy P and Farrés J, Methods and systems for identifying molecules or processes of biological interest by using knowledge discovery in biological data, 2010

9. Bray D, Bourret RB and Simon MI, Computer simulation of the phosphorylation cascade controlling bacterial chemotaxis, Mol Biol Cell 4(5): pp469-482, 1993

10. Terzopoulous, Artificial fishes: autonomous locomotion, perception, behavior, and learning in a simulated physical world, Artif Life 1: pp327-351, 2011

11. Sims, Evolving 3D morphology and behavior by competition, Artif Life 1: pp353-372, 1994

12. Tomita M, et al, E-CELL: software environment for whole-cell simulation, Bioinformatics 15(1): pp72-84, 1999

13. Pache RA et al, Towards a molecular characterisation of pathological pathways, FEBS Lett 582(8): pp1,259-1,265, 2008

Read full article from PDF >>

Rate this article You must be a member of the site to make a vote.  
Average rating:

There are no comments in regards to this article.

José Manuel Mas is Chief Operating Officer and Founder at Anaxomics Biotech, a company focused on identifying new potential indications for drugs by understanding the mechanisms of action (reprofiling). From 2008-2012 he held the position of European Union Head of the Collaborative Research Department at RPS Inc. He has a degree in Biochemistry, a Master’s degree in Biotechnology and a PhD in Biocomputing, specialising in protein structure, drug design and systems biology.

Albert Pujol Torras is currently working for IRB and Anaxomics for the Data Analysis department, where he is involved in the development of novel systems biology data mining technologies. Albert is an expert in computer vision, artificial intelligence and statistical pattern recognition. He has a degree in Computer Science and a PhD in Computer Science specialising in computer vision and face recognition.

José Manuel Mas
Albert Pujol Torras
Print this page
Send to a friend
Privacy statement
News and Press Releases

Automated Enzyme Analyzers Streamline Enzyme Assay Applications

SUNNYVALE, Calif. – (September 16, 2021) – Multi-enzyme analysis can now be achieved through a single, fully-automated enzyme analyzer, benefitting industrial enzyme manufacturers and enzyme users in the food, beverage, biopharmaceutical, fuel ethanol and chemical, and solid waste management industries.
More info >>

White Papers



©2000-2011 Samedan Ltd.
Add to favourites

Print this page

Send to a friend
Privacy statement