I have a variety of brain regions and information about how they're connected. (Not FMRI in this case, but the diffusion information about the white matter tracts connecting gray matter regions.)

I'd like to graph this information sensibly to deliver to the scientist who asked me for this information along with a spreadsheet describing each connection. I'd like to represent both the positions of the gray matter regions (I can project them onto the mid-sagittal plane to get 2D positions or do something similar) and the characteristics of the connections. (Probably just 2 or maybe 3, so I imagine color and thickness would cut it.)

I'm not very familiar with graph visualization tools -- I started with the graphviz dot format -- but now I'm reading about a multitude of tools. What sort of graph visualization tools ought I to use?


gephi is new, free and powerful. Recommended.

  • $\begingroup$ Exactly what I've been doing tutorials for for the last hour. It's pretty awesome. $\endgroup$ – Thomas Nov 8 '11 at 23:31

I thought I'd describe my experience with gephi.

I ended up writing some hacky code to write the gemf xml file myself because the pygexf python module seemed out of date. After a few tutorials, I created these visualizations: screenshot of use of gephi graph program output of gephi

Relative positions of nodes as projected onto the mid-sagittal plane of the brain were retained, with left and right hemispheres grouped together on the left and right. Edge thickness and color represent the quantities I'm most interested presenting.

Especially helpful in interpreting my data (where the edges are where all the interesting information is) was the ability to filter these edges based on arbitrary attributes.

I'm having trouble imagining a tool that would have been better, but these visualizations aren't great for presenting this edge information I'm interested in. I'm going to explore a little more, if I end up using gephi I'll accept JohnRos's answer and create a new question asking how I can improve these figures.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.