Network Diffusion

Website redesign inprogress, New design will be available on Sep 20,2019 !!

About Network Diffusion

Graph-based information diffusion (GID) method {PMID: 21179190 , 25126794} was applied to measure how well two groups of genes, Test & GS, are connected to each other in a protein-protein interaction network, STRING v10.0. The network is from string-db. We use the combined score that cover evidences from all sources. The network contains 19,247 genes and gene products. Through GID, functional information is propagated from genes of interest to all genes in the network through their connections. Genes receiving significantly more diffusion signals than random are more connected to and, thus functionally related to the original genes. If Test & GS overlap, diffusion signals will be initiated from 3 groups: Test only, GS only, and overlap. As diffusion is diffused from one group, the other groups will be validated whether they receive more signals than random.

We validate how well two gene groups are connected to each other by how well GID from one group recovers genes in another group through area under the curve (AUC) for receiver operating characteristic (ROC). For a diffusion experiment of each pair of groups, random genes are selected for the group where diffusion signals come from and for the group where signals come to. Random genes are randomly selected from genes in the network but not in Test & GS, and either with similar connectivity degrees with the group of interest or regardless of their degrees. Random is performed 100 times to obtain a distribution of random AUCs, which is tested for normality. Z-scores are computed for the experimental AUC based on the distribution of the random AUCs. Zf is the z-score when we compare experimental AUC with random AUCs of the genes where diffusion signals come “from”. Zt is the z-score comparing to random AUCs when we resample the genes that diffusion signals come “to”. Z-scores above 2.0 suggest that the pair of genes groups are significantly connected to each other.

Download content