Dear all, I'm using GOSemSim for computing semantic similarities (semsim) between clusters of genes. The Problem: I'm obtaining semsim==NA when comparing clusters of genes from which I actually expect a value (i.e. semsim != NA). However, if I compute semsim between individual genes (present in the analyzed clusters) I obtain semsim values != NA. For computing semsim between clusters, GOSemSim computes the pairwise semsim between the GO terms present in each of the clusters. Therefore I would expect that, if for at least one pair of genes (one from each cluster) is possible to compute a semsim != NA, then the semsim between the two clusters should also be different from NA. Is this situation a problem of GOSemSim or is there a wrong assumption from my side? Thanks a lot in advance for your time, Manuel PS: GOSemSim version: Ver. 1.8.3 (as from the DESCRIPTION file in the installation folder). This version is the one currently available in Bioconductor. source("http://bioconductor.org/biocLite.R") biocLite("GOSemSim") -- output of sessionInfo(): Code Examples: library("GOSemSim") library(GO.db) library(org.Hs.eg.db) clust_as <- c("10603","3551","8470","8915","5295","5518","2810","7534" ,"5591","2887","51177","57630") clust_r1 <- c("23761") clust_r2 <- c("30835") clust_r3 <- c("54332") # Comparing clusters > clusterSim(clust_as, clust_r1, ont="BP", organism="human", measure="Resnik", combine="rcmax.avg", drop="NULL") [1] NA > clusterSim(clust_as, clust_r2, ont="BP", organism="human", measure="Resnik", combine="rcmax.avg", drop="NULL") [1] NA > clusterSim(clust_as, clust_r3, ont="BP", organism="human", measure="Resnik", combine="rcmax.avg", drop="NULL") [1] NA # Comparing individual genes (the first one from each cluster) > geneSim("10603", "23761", ont="BP", organism="human", measure="Resnik", combine="rcmax.avg", drop="NULL") [1] 0.11 > geneSim("10603", "30835", ont="BP", organism="human", measure="Resnik", combine="rcmax.avg", drop="NULL") [1] 0.176 > geneSim("10603", "54332", ont="BP", organism="human", measure="Resnik", combine="rcmax.avg", drop="NULL") [1] 0.266 -- Sent via the guest posting facility at bioconductor.org.

Hi, As you reported that you are using GOSemSim version 1.8.3, that means you are using Bioconductor 2.7. The current release of Bioconductor is 2.10, please update your R and Bioconductor. Please read the vignette of GOSemSim before you using it, the clusterSim accept gene lists, not vectors. Best Regards! Guangchuang Yu --~--~---------~--~----~------------~-------~--~----~ Department of Biochemistry & Molecular Biology College of Life Science and Technology Jinan University, 601 Huangpu Ave. W. Guangzhou 510632, P.R. China Tel: +86-20-85222677 Email: guangchuangyu@gmail.com guangchuangyu@jnu.edu.cn Publications: http://bit.ly/IpzxpH -~----------~----~----~----~------~----~------~--~--- On Tue, Jul 24, 2012 at 4:08 PM, Manuel A.Alonso Tarajano [guest] < guest@bioconductor.org> wrote: > > Dear all, > > I'm using GOSemSim for computing semantic similarities (semsim) between > clusters of genes. > > The Problem: > I'm obtaining semsim==NA when comparing clusters of genes from which I > actually expect a value (i.e. semsim != NA). > > However, if I compute semsim between individual genes (present in the > analyzed clusters) I obtain semsim values != NA. > For computing semsim between clusters, GOSemSim computes the pairwise > semsim between the GO terms present in each of the clusters. Therefore I > would expect that, if for at least one pair of genes (one from each > cluster) is possible to compute a semsim != NA, then the semsim between the > two clusters should also be different from NA. > > Is this situation a problem of GOSemSim or is there a wrong assumption > from my side? > > Thanks a lot in advance for your time, > > Manuel > > PS: > GOSemSim version: > Ver. 1.8.3 (as from the DESCRIPTION file in the installation folder). > This version is the one currently available in Bioconductor. > source("http://bioconductor.org/biocLite.R") > biocLite("GOSemSim") > > > > -- output of sessionInfo(): > > Code Examples: > > library("GOSemSim") > library(GO.db) > library(org.Hs.eg.db) > > clust_as <- > c("10603","3551","8470","8915","5295","5518","2810","7534","5591","2 887","51177","57630") > clust_r1 <- c("23761") > clust_r2 <- c("30835") > clust_r3 <- c("54332") > > # Comparing clusters > > clusterSim(clust_as, clust_r1, ont="BP", organism="human", > measure="Resnik", combine="rcmax.avg", drop="NULL") > [1] NA > > clusterSim(clust_as, clust_r2, ont="BP", organism="human", > measure="Resnik", combine="rcmax.avg", drop="NULL") > [1] NA > > clusterSim(clust_as, clust_r3, ont="BP", organism="human", > measure="Resnik", combine="rcmax.avg", drop="NULL") > [1] NA > > # Comparing individual genes (the first one from each cluster) > > geneSim("10603", "23761", ont="BP", organism="human", measure="Resnik", > combine="rcmax.avg", drop="NULL") > [1] 0.11 > > geneSim("10603", "30835", ont="BP", organism="human", measure="Resnik", > combine="rcmax.avg", drop="NULL") > [1] 0.176 > > geneSim("10603", "54332", ont="BP", organism="human", measure="Resnik", > combine="rcmax.avg", drop="NULL") > [1] 0.266 > > > -- > Sent via the guest posting facility at bioconductor.org. > [[alternative HTML version deleted]]