CS446/546 - Class Session 13 - Similarity and Hierarchical Clustering

In this class session we are going to hierachically cluster (based on Sorensen-Dice similarity) vertices in a directed graph from a landmark paper on human gene regulation (Neph et al., Cell, volume 150, pages 1274-1286, 2012; see PDF on Canvas)

Let's start by having a look at the Neph et al. data file, neph_gene_network.txt. It is in edge-list format, with no header and no "interaction" column. Just two columns, first column contains the "regulator gene" and the second column contains the "target gene":

head neph_gene_network.txt
AHR     BCL6
AHR     BHLHE41
AHR     BPTF
AHR     CEBPA
AHR     CNOT3
AHR     CREB1

Now let's load the packages that we will need for this exercise



In [19]:

    
suppressPackageStartupMessages(library(igraph))

Using read.table, read the file shared/neph_gene_network.txt; name the two columns of the resulting data frame, regulator and target. Since there is no header, we will use header=FALSE:



In [21]:

    
edge_list_neph <- read.table("shared/neph_gene_network.txt",
                            header=FALSE,
                            sep="\t",
                            stringsAsFactors=FALSE,
                            col.names=c("regulator","target"))

Load the edge-list data into a Graph object in igraph, using graph_from_data_frame. Make the graph undirected



In [22]:

    
neph_graph <- graph_from_data_frame(edge_list_neph, directed=FALSE)
summary(neph_graph)









    



IGRAPH c6d5c31 UN-- 538 47945 -- 
+ attr: name (v/c)

Get the adjacency matrix for the graph, using get.adjacency, and assign to matrix g_matrix



In [23]:

    
g_matrix <- get.adjacency(neph_graph)

Compute the Sorensen-Dice similarity of this graph, and assign to similarity matrix S, using the igraph similarity function and specifying method="dice"



In [25]:

    
S = similarity(neph_graph, method="dice")

Compute a distance matrix by calling as.dist on the object 1-S; assign to object D



In [26]:

    
D = as.dist(1 - S)

Perform average-linkage hierarchical clustering on this distance object, using hclust with method=average



In [27]:

    
cc = hclust(D, method="average")

Plot the dendrogram using plot and make the labels really small using cex=0.01



In [28]:

    
plot(cc, cex=0.01)



In [ ]: