Vinay Vinay - 1 month ago 12
R Question

Creating Subgraph using igraph in R

I need to obtain a subgraph of the seed nodes (the input list of nodes; file.txt) and their first interactors (neighbours) from a graph (g) using igraph. Unfortunately, I am ending up with only a single node in the subgraph, and not all the rest of the nodes and edges (vertices) which interlink them.

g<-read.graph("DATABASE.ncol",format="ncol",directed=FALSE) #load the data
g2<-simplify(g, remove.multiple=TRUE, remove.loops=TRUE) # Remove the self-loops in the data
DAT1 <- readLines("file.txt") #It provides a character vector right away
list_nodes_1 = neighbors(g2, DAT1) #list of nodes to be fetched in subnetwork
list_nodes_1 # 16
g3 <- induced.subgraph(graph=g2,vids=DAT1) #subnetwork construction
g3 # GRAPH UN-- 1 0 --; indicating only one node
plot (g3)


Any suggestions for obtaining the whole subnetwork (including nodes and vertices)? or is there any other function available for creating sub-graphs?

DATABASE.ncol:

MAP2K4 FLNC
MYPN ACTN2
ACVR1 FNTA
GATA2 PML
RPA2 STAT3
ARF1 GGA3
ARF3 ARFIP2
ARF3 ARFIP1
XRN1 ALDOA
APP APPBP2
APLP1 DAB1
CITED2 TFAP2A
EP300 TFAP2A
APOB MTTP
ARRB2 RALGDS
CSF1R GRB2
PRRC2A GRB2
LSM1 NARS
SLC4A1 SLC4A1AP
BCL3 BARD1


It is a simple text file with one edge per line. An edge is defined by two symbolic vertex names delimited by tab:

file.txt

ALDOA
APLP1
GRB2
RPA2
FLNC
BCL3
APP
RALGDS
PRRC2A
NARS
LSM1
GGA3
FNTA

Answer

I'm not sure to have completely understood your problem, so I have created a (hopefully) self-explanatory example:

# for example reproducibility
set.seed(123)

# create a fake undirected graph
D <- read.table(
sep=',',
header=T,
text=
'from,to
A,B
A,C
D,E
F,G
H,I')

g1 <- graph.data.frame(D,directed=F)
plot(g1)

# we want a sub-network containing the floowing nodes:
subv <- c('A','B','H')

# first method: 
# create a sub-network composed by ONLY the nodes in subv and the edges 
# between them
g2 <- induced.subgraph(graph=g1,vids=subv)
plot(g2)

# second method: 
# create a sub-network composed by the nodes in subv and, if some of them is
# connected to other nodes (even if not in subv), take also them 
# (and of course include all the edges among this bunch of nodes). 
sg1 <- decompose.graph(g1,mode="weak")
neighverts <- unique(unlist(sapply(sg1,FUN=function(s){if(any(V(s)$name %in% subv)) V(s)$name else NULL})))
g3 <- induced.subgraph(graph=g1,vids=neighverts)
plot(g3)

Graph g1 :

g1

Graph g2 :

g2

Graph g3 :

g3