Download Network

Manually Generated Interaction Networks

IAMBEE needs a network scaffold to drive its analysis. Precompiled networks are provided for Escherichia coli and Salmonella typhimurium.

The user can also download his own network from STRING. STRING database contains interactions between genes for both model and non-model organisms. Interactions are derived from curated databases, literature, co-expression and co-occurrence information using a Bayesian integration scheme. STRING provides for each interaction a confidence level. IAMBEE works the best if the network is not too overconnected. Including too many edges might make computations prohibitively long (the number of paths increases with the network size). In addition including too many false positive edges might result in spurious predictions. We advise the user to start with a high confidence network (edge confidence > 0.9)

Prior to running IAMBEE make sure the genes or pathways you assume relevant in the process you study are in the network. If not, but you have prior evidence they are connected in the network (expert knowledge), add them manually to the network file. It will make the network scaffold much more informative for analyzing your experimental dataset.

The network downloaded from STRING only contains edges of the pp type. This is because STRING is developed to predict functional interactions that are a proxy of predominantly protein protein interactions.

Salmonella typhymurium LT2

The Salmonella typhymurium LT2 interaction network is constructed using KEGG V77.0 for metabolic and signalling interactions, the STRING interaction database V10 for physical protein-protein interactions and the protein-DNA and sRNA interactions converted from RegulonDB.

The network file contains 11 interaction types (header) and 13104 connections between genes (body).

Escherichia coli K-12

This interaction network is constructed using data from the KEGG interaction database for metabolic and signalling interactions, the STRING database for protein-protein interactions and the RegulonDB for gene regulatory interactions such as protein-DNA and sRNA interactions. This network is an updated version from the original PheNetic publication (De Maeyer, 2013).

The interaction network contains 6 interaction types (header) and 18631 connections between genes (body).