03 September 2015 0 994 Report

We constructed a protein-protein interaction network with some genes using Cytoscape version 3.1.0, Cluepedia-Action-STRING v10-2015. We noticed that there are not some interactions between some of genes in spite of existing evidence, for example Vegfa is one of our genes but in the constructed PPI network, it has not any edges, however we know it has an interaction with Hif1a (another our genes). Again, we constructed the network by Cytoscape version 3.2.1, Action-STRING v10-2015 but the result did not change. We used STRING database directly and the PPI network in the STRING was different than the PPI network of Cytoscape. The Vegfa was a deeply connected node in the network as we expected. We changed the confidence threshold of STRING from Low confidence to High confidence to see the state of Vegfa in the network. In all changes, this gene was involved in the network.

Then, we constructed the PPI network of our genes by Cytoscape version 3.1.0, Clupedia-Action-STRING v9-2012. The Vegfa showed interactions with other genes such as Hif1a. However this network had fewer edges (as you know the older version (v9-2012) does not have some edges such as inhibition).

We are confused because the newer version of Action-STRING has more edges than older version but it misses some obvious edges like Vegfa and Hif1a

 Could you help us to find better way to construct the network?

Do you think why the two versions of Cluepedia-Action-STRING are different?

Best Regards

Maryam

More Maryam Abedi's questions See All
Similar questions and discussions