Hi list, I am working with a network, in a table form like this: Gene1 Gene5 0.76 Gene1 Gene79 0.89 Gene2 Gene8 0.97 ... Basically, each row represents two genes that interact, and their pearson correlation coefficient (from 0.7 to 1). What I want to get is a plot representing how the number of nodes and edges (separately) decrease when the pearson correlation cut off increases (from 0.7 to 1). What I have in mind is creating a loop and generate values for the plot, but I have never done anything similar in R and I am not quite sure how to do it or it that would work. Any ideas? Help is much appreciated, thanks, -- Hernando Martínez Vergara [[alternative HTML version deleted]]

On Thu, Jul 1, 2010 at 4:18 AM, Hernando Martínez <hernybiotec@gmail.com>wrote: > Hi list, > I am working with a network, in a table form like this: > > Gene1 Gene5 0.76 > Gene1 Gene79 0.89 > Gene2 Gene8 0.97 > ... > > Basically, each row represents two genes that interact, and their pearson > correlation coefficient (from 0.7 to 1). > What I want to get is a plot representing how the number of nodes and edges > (separately) decrease when the pearson correlation cut off increases (from > 0.7 to 1). > What I have in mind is creating a loop and generate values for the plot, > but > I have never done anything similar in R and I am not quite sure how to do > it > or it that would work. > Any ideas? Help is much appreciated, thanks, > > Hi, Hernando. You might want to look at the ecdf function given the problem you describe. Sean [[alternative HTML version deleted]]

You just need to build a matrix with two rows (one for nodes and one for edges), and as many columns as correlations cutoffs you want (e.g. if you have 0.7, 0.8, 0.9 and 1 it will have 4 columns). Then you just need to do a barplot for that matrix. As an example, if you have 4 correlation cutoffs: a<-matrix(8:1,2,4) #Matrix with number of nodes and edges for each PCC barplot(a,beside=TRUE,names.arg=c("0.7","0.8","0.9","1"),xlab="PCC threshold",ylab="Network parameters",col=c("blue","gray")) legend("topright",legend=c("nodes","edges"),fill=c("blue","gray")) To build the matrix there are some alternatives, and certainly one of them is a loop. Besides that, this is more an R question than a Bioconductor one, so maybe you should consider posting this to the R list. Best, Xavi. -----Mensaje original----- De: bioconductor-bounces at stat.math.ethz.ch [mailto:bioconductor- bounces at stat.math.ethz.ch] En nombre de Hernando Mart?nez Enviado el: jueves, 01 de julio de 2010 10:19 Para: Bioconductor at stat.math.ethz.ch Asunto: [BioC] Network plot Hi list, I am working with a network, in a table form like this: Gene1 Gene5 0.76 Gene1 Gene79 0.89 Gene2 Gene8 0.97 ... Basically, each row represents two genes that interact, and their pearson correlation coefficient (from 0.7 to 1). What I want to get is a plot representing how the number of nodes and edges (separately) decrease when the pearson correlation cut off increases (from 0.7 to 1). What I have in mind is creating a loop and generate values for the plot, but I have never done anything similar in R and I am not quite sure how to do it or it that would work. Any ideas? Help is much appreciated, thanks, -- Hernando Mart?nez Vergara [[alternative HTML version deleted]]