Limits...
Metabolic tinker: an online tool for guiding the design of synthetic metabolic pathways.

McClymont K, Soyer OS - Nucleic Acids Res. (2013)

Bottom Line: The fast pace of developments in molecular biology increasingly makes it possible to experimentally redesign existing pathways and implement de novo ones in microbes or using in vitro platforms.For such experimental studies, the bottleneck is shifting from implementation of pathways towards their initial design.Here, we present an online tool called 'Metabolic Tinker', which aims to guide the design of synthetic metabolic pathways between any two desired compounds.

View Article: PubMed Central - PubMed

Affiliation: Computer Science, College of Engineering, Mathematics, and Physical Sciences, University of Exeter, Exeter EX4 4QF, UK.

ABSTRACT
One of the primary aims of synthetic biology is to (re)design metabolic pathways towards the production of desired chemicals. The fast pace of developments in molecular biology increasingly makes it possible to experimentally redesign existing pathways and implement de novo ones in microbes or using in vitro platforms. For such experimental studies, the bottleneck is shifting from implementation of pathways towards their initial design. Here, we present an online tool called 'Metabolic Tinker', which aims to guide the design of synthetic metabolic pathways between any two desired compounds. Given two user-defined 'target' and 'source' compounds, Metabolic Tinker searches for thermodynamically feasible paths in the entire known metabolic universe using a tailored heuristic search strategy. Compared with similar graph-based search tools, Metabolic Tinker returns a larger number of possible paths owing to its broad search base and fast heuristic, and provides for the first time thermodynamic feasibility information for the discovered paths. Metabolic Tinker is available as a web service at http://osslab.ex.ac.uk/tinker.aspx. The same website also provides the source code for Metabolic Tinker, allowing it to be developed further or run on personal machines for specific applications.

Show MeSH
A graph representation of the URN to illustrate the scale and density of this directional graph. Nodes represent compounds, and edges represent reactions between these compounds. The more connected compounds are clustered in the centre of the graph and the lesser connected nodes at the extremities. The connectivity distribution of the nodes is shown in a histogram in the inset, which can be fitted to a power law distribution with an exponent of 3.4.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3675468&req=5

gkt234-F1: A graph representation of the URN to illustrate the scale and density of this directional graph. Nodes represent compounds, and edges represent reactions between these compounds. The more connected compounds are clustered in the centre of the graph and the lesser connected nodes at the extremities. The connectivity distribution of the nodes is shown in a histogram in the inset, which can be fitted to a power law distribution with an exponent of 3.4.

Mentions: The URN that Tinker searches is produced from a periodically updated internal database of metabolic compounds and reactions available from the CHEBI (9) and Rhea (8) databases. At the time of writing, this internal database contains ∼20 000 reactions and 29 000 compounds, which have been manually qualified against references to the literature. The Rhea database provides a large set of curated metabolic reactions, which in many cases are annotated with the known direction of the reaction. Using this information a partially directed network (i.e. the URN) is produced, where reactions with missing direction information is completed using a prediction of the reaction thermodynamics, described in the next section below. The URN is shown in Figure 1. Similar to networks created from metabolic data of single species (20), the URN has a connectivity distribution that could be described by a power law with an average degree of 3.2 and peak degree of 4.9. Thus, the URN contains certain hubs, i.e. highly connected compounds such as cofactors ADP and NADH+. For the purposes of Tinker, the inclusion of these compounds in the pathway search is not feasible, as their usage as intermediary compounds in possible synthetic paths is not realistic and their presence exponentially increases the search time. With these considerations, we exclude the most highly connected compounds (those with a degree >650) from the search process, and also excluded a number of well-known cofactors (such as dATP) from being included as primary compounds in the pathways. It should be noted that these compounds can still be given as source and target compounds. The full list of excluded compounds is given in Table 1 and provided on the Tinker website and can be altered in the source code of the program.Figure 1.


Metabolic tinker: an online tool for guiding the design of synthetic metabolic pathways.

McClymont K, Soyer OS - Nucleic Acids Res. (2013)

A graph representation of the URN to illustrate the scale and density of this directional graph. Nodes represent compounds, and edges represent reactions between these compounds. The more connected compounds are clustered in the centre of the graph and the lesser connected nodes at the extremities. The connectivity distribution of the nodes is shown in a histogram in the inset, which can be fitted to a power law distribution with an exponent of 3.4.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3675468&req=5

gkt234-F1: A graph representation of the URN to illustrate the scale and density of this directional graph. Nodes represent compounds, and edges represent reactions between these compounds. The more connected compounds are clustered in the centre of the graph and the lesser connected nodes at the extremities. The connectivity distribution of the nodes is shown in a histogram in the inset, which can be fitted to a power law distribution with an exponent of 3.4.
Mentions: The URN that Tinker searches is produced from a periodically updated internal database of metabolic compounds and reactions available from the CHEBI (9) and Rhea (8) databases. At the time of writing, this internal database contains ∼20 000 reactions and 29 000 compounds, which have been manually qualified against references to the literature. The Rhea database provides a large set of curated metabolic reactions, which in many cases are annotated with the known direction of the reaction. Using this information a partially directed network (i.e. the URN) is produced, where reactions with missing direction information is completed using a prediction of the reaction thermodynamics, described in the next section below. The URN is shown in Figure 1. Similar to networks created from metabolic data of single species (20), the URN has a connectivity distribution that could be described by a power law with an average degree of 3.2 and peak degree of 4.9. Thus, the URN contains certain hubs, i.e. highly connected compounds such as cofactors ADP and NADH+. For the purposes of Tinker, the inclusion of these compounds in the pathway search is not feasible, as their usage as intermediary compounds in possible synthetic paths is not realistic and their presence exponentially increases the search time. With these considerations, we exclude the most highly connected compounds (those with a degree >650) from the search process, and also excluded a number of well-known cofactors (such as dATP) from being included as primary compounds in the pathways. It should be noted that these compounds can still be given as source and target compounds. The full list of excluded compounds is given in Table 1 and provided on the Tinker website and can be altered in the source code of the program.Figure 1.

Bottom Line: The fast pace of developments in molecular biology increasingly makes it possible to experimentally redesign existing pathways and implement de novo ones in microbes or using in vitro platforms.For such experimental studies, the bottleneck is shifting from implementation of pathways towards their initial design.Here, we present an online tool called 'Metabolic Tinker', which aims to guide the design of synthetic metabolic pathways between any two desired compounds.

View Article: PubMed Central - PubMed

Affiliation: Computer Science, College of Engineering, Mathematics, and Physical Sciences, University of Exeter, Exeter EX4 4QF, UK.

ABSTRACT
One of the primary aims of synthetic biology is to (re)design metabolic pathways towards the production of desired chemicals. The fast pace of developments in molecular biology increasingly makes it possible to experimentally redesign existing pathways and implement de novo ones in microbes or using in vitro platforms. For such experimental studies, the bottleneck is shifting from implementation of pathways towards their initial design. Here, we present an online tool called 'Metabolic Tinker', which aims to guide the design of synthetic metabolic pathways between any two desired compounds. Given two user-defined 'target' and 'source' compounds, Metabolic Tinker searches for thermodynamically feasible paths in the entire known metabolic universe using a tailored heuristic search strategy. Compared with similar graph-based search tools, Metabolic Tinker returns a larger number of possible paths owing to its broad search base and fast heuristic, and provides for the first time thermodynamic feasibility information for the discovered paths. Metabolic Tinker is available as a web service at http://osslab.ex.ac.uk/tinker.aspx. The same website also provides the source code for Metabolic Tinker, allowing it to be developed further or run on personal machines for specific applications.

Show MeSH