Elhanan Borenstein, Ph.D.

NetSeed>Examples
Here we provide some example networks that can be used to try out NetSeed and to learn the basic concepts of the "seed set" of a network.

The example networks are available to download as tab-separated text files or SBML files.


Example 1: A simple network.

In this example, we have a simple network with one seed connected to a loop with three nodes.

Activating node 1 would allow you to activate nodes 2, 3, and 4, but no node can activate node 1, so node 1 is a seed. For a metabolic network, node 1 would represent a metabolite obtained from the environment.

The NetSeed>Web application will optionally use Cytoscape Web to draw the network. It presents seeds in red, and the non-seed nodes in orange.

Download the example file as a text file (tab-separated) or SBML file.

Example 2: Seed groups and seed confidence levels.

In this example, we have a network with four seeds.

Node A is a seed, as it cannot be activated by any other node in the network.

In contrast, nodes F, G, and H are all marked as seeds, but can be seen to be interdependent: Activating one of these nodes would activate the rest, but at least one must be active to activate the rest. These nodes form a "seed group".

To quantify the relevance of each identified seed, we assign each seed a "confidence level", ranging from 0 to 1. A confidence level of 0 would correspond to a non-seed node, while a 1 would correspond to a seed that cannot be activated by another node. Seeds which belong to a seed group with more than 1 seed are given a fractional confidence level, the inverse of the number of seeds in the group. Nodes F, G, and H would then each have a confidence level of 1/3.

Download the example file as a text file (tab-separated) or SBML file.

Example 3: Networks with more than one component.

In this example, we have a network with more than one component. A component is a connected set of nodes.

In NetSeed, you have the ability to restrict the calculation of the seed set to only the giant component (the largest connected component), or connected components having at least a minimum number of nodes. This is useful for calculating the seed sets of metabolic networks created from genomic data, as there are often small, trivial components due to incomplete annotation of the genes.

If you choose to restrict the calculation to a subset of nodes, NetSeed>Web will color the nodes that were not considered (the "ignored nodes") gray. In this figure, all the components with fewer than 4 nodes are colored gray.

Download the example file as a text file (tab-separated) or SBML file.

Example 4: The metabolic network of Buchnera aphidicola.

For our last example, we present the metabolic network of the obligate endocellular symbiont Buchnera aphidicola. The seed set of the metabolic network of this organism was previously published in Borenstein et al. (2008) (supporting information), and in this example, we investigate the giant component of the network.

Note that in this example, the node names are given by the KEGG compound IDs, which NetSeed can use to create links to the KEGG compound definitions. Try out the interactive example to explore the seed set of the metabolic network of Buchnera aphidicola.

Download the example file as a text file (tab-separated).