This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Gene-Gene interaction Network Analysis and Visualization Software

I searched this forum and found many related questions about network. I will try to be as precise as necessary.

Suppose I get a partial correlation (PC) matrix (from RNA-seq) and want to use the PC matrix to build my gene-gene interaction (GGI) network. The edges can be pruned/trimmed by a threshold PC value. Now I want to

  • Visualize the vertices and edges

  • Display + and - PC values differently (e.g. red for + & blue for -)

  • Visualize the magnitude of the PC (e.g. a thicker line for a bigger PC and a thinner line for a smaller PC)

  • Ideally a R package, but a java or perl package or anything that runs on a ubuntu machine would work

BTW, is there a good software product that can generate directional graph (interactions with directions)? I know a set of time series data would be needed to generate a directional graph. Time series is often not satisfied, but I guess in general a set of samples treated with increasing concentrations of a compound with the same length of time might simulate a time series.

Thanks in advance!

r rna-seq gene
• 4,611 views • link
updated by Jean-Karim Heriche Jean-Karim Heriche • written by moxu moxu

2 answers

Without having tried it, I think Cytoscape with the ARACNE app should allow you to do everything you ask for.

• 0 views • link
written by Lars Juhl Jensen Lars Juhl Jensen

Thanks for the reply. I will study more on Cytoscape and ARACNE.

• 0 views • link
written by moxu moxu

The already mentioned Cytoscape is a standard for biological network visualization but it is meant to run as a stand-alone application. Using R, there's the igraph package and a good tutorial on network visualization with igraph. For perl, there's the GraphViz2 module.

• 0 views • link
written by Jean-Karim Heriche Jean-Karim Heriche

Thanks much for the reply, very useful information.

• 0 views • link
written by moxu moxu

I believe this is normal behaviour. When reading from a matrix, the nodes are derived from the row/column names. You can easily remove isolated nodes by filtering based on degree. The behaviour may also depend on the mode, weighted and diag parameters of the from_adjacency() function.

• 0 views • link
written by Jean-Karim Heriche Jean-Karim Heriche

Log in to answer this question.

More posts like this