Dear All, I am doing my analysis for time series data and have to perform template matching. I am new to this concept and need some help to proceed further Here are the details of my analysis: I have 7000x 82 matrix in which rows are the genes and col are samples respectively. I have to define templates using 6 different combination for which have used the following code; templates <- permutations(3,6,c(-1,0,1),repeats=T) comb1<-c (1,2,3,4) com2 <- c (5,6,7,8) com3 <- c (9,10,11,12) com4 <- c(13,14,15,16) com5 <- c(17,18,19,20) com6 <- c(21,22,23,24) com7 <- c(25,26,27,28) To get the 729 profiles for the templates using permutation code is as follows profile <- matrix(0,nrow=729,ncol=82) for (i in 1:729){ profile[i,hf1hr.cols] <- templates[i,1] profile[i,hf6hr.cols] <- templates[i,2] profile[i,hf24hr.cols] <-templates[i,3] profile[i,cho1hr.cols] <- templates[i,4] profile[i,cho6hr.cols] <- templates[i,5] profile[i,cho24hr.cols] <- templates[i,6] } once i run the above code I have 729 x82 profiles. I have to calculate correlation of each gene with each profile. Should have a 7000 X 82 size matrix. Then have to find for each gene,the profile with maximum absolute correlation and use that number for that gene.I should end up with 7000 X 1 sized profile assignment. I am struck in between my analysis please help with this!!!!!!!!!!!!!!!!!!!! It would be greatly appreciated if I get some suggestions,code,or links. Have a great day -- Thanks Harikrishnadhar Sonnenahalli [[alternative HTML version deleted]]

---------- Forwarded message ---------- From: Harikrishnadhar <hari.bombex@gmail.com> Date: Mon, Aug 30, 2010 at 3:36 PM Subject: template matching To: Bioconductor@stat.math.ethz.ch Dear All, I am doing my analysis for time series data and have to perform template matching. I am new to this concept and need some help to proceed further Here are the details of my analysis: I have 7000x 82 matrix in which rows are the genes and col are samples respectively. I have to define templates using 6 different combination for which have used the following code; templates <- permutations(3,6,c(-1,0,1),repeats=T) comb1<-c (1,2,3,4) com2 <- c (5,6,7,8) com3 <- c (9,10,11,12) com4 <- c(13,14,15,16) com5 <- c(17,18,19,20) com6 <- c(21,22,23,24) com7 <- c(25,26,27,28) To get the 729 profiles for the templates using permutation code is as follows profile <- matrix(0,nrow=729,ncol=82) for (i in 1:729){ profile[i,hf1hr.cols] <- templates[i,1] profile[i,hf6hr.cols] <- templates[i,2] profile[i,hf24hr.cols] <-templates[i,3] profile[i,cho1hr.cols] <- templates[i,4] profile[i,cho6hr.cols] <- templates[i,5] profile[i,cho24hr.cols] <- templates[i,6] } once i run the above code I have 729 x82 profiles. I have to calculate correlation of each gene with each profile. Should have a 7000 X 729 size matrix. Then have to find for each gene,the profile with maximum absolute correlation and use that number for that gene.I should end up with 7000 X 1 sized profile assignment. I am struck in between my analysis please help with this!!!!!!!!!!!!!!!!!!!! It would be greatly appreciated if I get some suggestions,code,or links. Have a great day -- Thanks Harikrishnadhar Sonnenahalli -- Thanks Harikrishnadhar Sonnenahalli [[alternative HTML version deleted]]