I have a gene network that I clustered into 8 parts. I calculated the degree distribution of the clusters using Cytoscape and the slope of the clusters range from $0.7$ to about $1.3$.

From literature (reference) usual scale free networks (real world networks) have a slope range of $2>γ>3$. What could be a reason for my network to have the smaller slope value?

Edit: The number of nodes of each cluster ranges from 431 to 187 and the number of edges range from 1009 to 263. The network is bipartite in nature. The clusters are all very sparse which can be understood by the number of edges in the largest and smallest cluster. Are real world biological networks not always scale free. Can anyone point me towards publications that mention this fact. I am just confused about why my network doesn't conform to the norm but maybe it doesn't need to.

  • 1
    $\begingroup$ Could you add more information about your network? How many nodes does it have? How are them distributed in your 8 clusters? What kind of data are you using to create it (It seems like gene expression, but I'm not sure) ? $\endgroup$
    – llrs
    Jun 19, 2018 at 7:21
  • 1
    $\begingroup$ Yes, please edit and explain more about your network. 8 clusters suggests your network is likely very small. Is it also sparse, as biological networks often are? You also seem to be starting with the assumption that your network is scale free. Is it? It might not be, after all. $\endgroup$
    – terdon
    Jun 19, 2018 at 9:09
  • 1
    $\begingroup$ Actual networks are rarely scale free. $\endgroup$
    – Devon Ryan
    Jun 26, 2018 at 13:58
  • 1
    $\begingroup$ @Devon, this seems to be the answer rather a comment. Thanks for providing the reference (I lost it/I didn't save it when I read it). @ The Last Word: Despite not being a scale-free you can still work with your clusters, but take note that the topology of your network is not to be trusted. $\endgroup$
    – llrs
    Jun 26, 2018 at 14:26
  • 1
    $\begingroup$ @Devon I think there was another paper about networks in biology, but in general it seems like people is using topologies analysis assuming too many things without (first) checking them. $\endgroup$
    – llrs
    Jun 26, 2018 at 14:51


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.