I would like to normlization populus RNA seq with package CQN. I follow the code as show in manual, but after run i got the RPKM as minus(see detail below). What is the meaning of minus in RPKM? Or i done something wrong in my code? I try to visulize with cqn plot, and i think my data are not bias by GC content or lenght. I attached file of cqn plot for your information. Do you think i need to use your method to normailze data? Sincerely yours, Sermsawat T. #####normalize by CQN > #find GC content from phytozome populus genome > library(Rsamtools) > seq <- scanFa("pop.fa") > alph <- alphabetFrequency(seq, as.prob=TRUE) > gc <- rowSums(alph[,c("G", "C")]) > > library(cqn) > library(scales) > raw <- read.table("raw.txt", header=FALSE) > cqn.raw <- cqn(raw, lengths = width(seq), x = gc, sizeFactors = colSums(raw), verbose = TRUE) RQ fit .................. SQN . > > #cqn plot > par(mfrow=c(1,2)) > cqnplot(cqn.raw, n = 1, xlab = "GC content", lty = 1, ylim = c(1,7)) > cqnplot(cqn.raw, n = 2, xlab = "length", lty = 1, ylim = c(1,7)) > > #normalizedvalues > RPKM.cqn <- cqn.raw$y + cqn.raw$offset > head(RPKM.cqn) V1 V2 V3 V4 V5 V6 [1,] -2.3157759 -0.05011048 -1.5628979 -2.792141 -1.5042294 -2.1660771 [2,] 3.3638330 3.46092091 3.5447488 3.333357 3.1484934 3.4531245 [3,] -0.1924864 0.38536951 0.8976638 -1.134989 -0.1924864 0.4854206 [4,] -2.2808497 -0.50606822 -1.5613805 -2.802241 -1.4820398 -3.1508760 [5,] -2.2813593 -0.50602611 -1.5615417 -2.802341 -1.4822924 -3.1510596 [6,] 4.6875943 4.68565806 2.6239807 3.531895 3.5440903 1.9465045 V7 V8 V9 V10 V11 V12 [1,] -3.417837 -0.05011048 -0.05011048 -1.203393 -0.5058032 -1.4579155 [2,] 2.900329 3.07780614 3.41975117 3.693540 3.3387816 3.3901923 [3,] -1.369314 -0.05011048 0.58067142 0.437804 1.0192374 0.8445695 [4,] -3.459816 -0.38013895 -0.05011048 -1.182797 -0.5064885 -1.4613012 [5,] -3.459801 -0.38013895 -0.05011048 -1.183303 -0.5066405 0.1602583 [6,] 2.282431 2.19639308 1.39946079 4.753675 4.8337840 2.7983964 V13 V14 V15 V16 V17 V18 [1,] -4.303745 -1.844632 -3.1044275 -3.4759833 -0.1924864 -1.2015133 [2,] 3.116199 2.942757 3.2217711 2.9451114 2.8716179 3.4992808 [3,] -1.627192 -1.779755 -1.5114489 -0.8413003 -0.1924864 0.3091839 [4,] -3.289903 -2.870566 -3.1055409 -3.5218687 -0.3801390 -1.2399416 [5,] -4.290241 -2.870453 -2.1055097 -3.5218168 -0.3801390 -1.2400014 [6,] 3.685705 3.460004 0.8445695 2.4490306 2.6279081 0.5358322 > sessionInfo() R version 2.15.1 (2012-06-22) Platform: x86_64-pc-mingw32/x64 (64-bit) locale: [1] LC_COLLATE=English_United States.1252 [2] LC_CTYPE=English_United States.1252 [3] LC_MONETARY=English_United States.1252 [4] LC_NUMERIC=C [5] LC_TIME=English_United States.1252 attached base packages: [1] splines stats graphics grDevices utils datasets methods [8] base other attached packages: [1] scales_0.2.1 Rsamtools_1.8.5 Biostrings_2.24.1 [4] GenomicRanges_1.8.7 IRanges_1.14.4 BiocGenerics_0.2.0 [7] cqn_1.2.0 quantreg_4.81 SparseM_0.96 [10] preprocessCore_1.18.0 nor1mix_1.1-3 mclust_3.5 loaded via a namespace (and not attached): [1] bitops_1.0-4.1 colorspace_1.1-1 dichromat_1.2-4 labeling_0.1 [5] munsell_0.3 plyr_1.7.1 RColorBrewer_1.0-5 stats4_2.15.1 [9] stringr_0.6 zlibbioc_1.2.0 -------------- next part -------------- A non-text attachment was scrubbed... Name: cqnplot.pdf Type: application/pdf Size: 18583 bytes Desc: not available URL: <https: stat.ethz.ch="" pipermail="" bioconductor="" attachments="" 20120717="" 0b43c6ee="" attachment.pdf="">

I have been away. The rpkm values are on log2 scale, hence the negative values. Your plot shows, in my opinion, a large gc bias effect. Kasper On Tuesday, July 17, 2012, Sermsawat Tunlaya-Anukit wrote: > I would like to normlization populus RNA seq with package CQN. I follow the > code as show in manual, but after run i got the RPKM as minus(see detail > below). What is the meaning of minus in RPKM? Or i done something wrong in > my code? I try to visulize with cqn plot, and i think my data are not bias > by GC content or lenght. I attached file of cqn plot for your information. > Do you think i need to use your method to normailze data? > > Sincerely yours, > Sermsawat T. > > #####normalize by CQN > > #find GC content from phytozome populus genome > > library(Rsamtools) > > seq <- scanFa("pop.fa") > > alph <- alphabetFrequency(seq, as.prob=TRUE) > > gc <- rowSums(alph[,c("G", "C")]) > > > > library(cqn) > > library(scales) > > raw <- read.table("raw.txt", header=FALSE) > > cqn.raw <- cqn(raw, lengths = width(seq), x = gc, sizeFactors = > colSums(raw), verbose = TRUE) > RQ fit .................. > SQN . > > > > #cqn plot > > par(mfrow=c(1,2)) > > cqnplot(cqn.raw, n = 1, xlab = "GC content", lty = 1, ylim = c(1,7)) > > cqnplot(cqn.raw, n = 2, xlab = "length", lty = 1, ylim = c(1,7)) > > > > #normalizedvalues > > RPKM.cqn <- cqn.raw$y + cqn.raw$offset > > head(RPKM.cqn) > V1 V2 V3 V4 V5 V6 > [1,] -2.3157759 -0.05011048 -1.5628979 -2.792141 -1.5042294 -2.1660771 > [2,] 3.3638330 3.46092091 3.5447488 3.333357 3.1484934 3.4531245 > [3,] -0.1924864 0.38536951 0.8976638 -1.134989 -0.1924864 0.4854206 > [4,] -2.2808497 -0.50606822 -1.5613805 -2.802241 -1.4820398 -3.1508760 > [5,] -2.2813593 -0.50602611 -1.5615417 -2.802341 -1.4822924 -3.1510596 > [6,] 4.6875943 4.68565806 2.6239807 3.531895 3.5440903 1.9465045 > V7 V8 V9 V10 V11 V12 > [1,] -3.417837 -0.05011048 -0.05011048 -1.203393 -0.5058032 -1.4579155 > [2,] 2.900329 3.07780614 3.41975117 3.693540 3.3387816 3.3901923 > [3,] -1.369314 -0.05011048 0.58067142 0.437804 1.0192374 0.8445695 > [4,] -3.459816 -0.38013895 -0.05011048 -1.182797 -0.5064885 -1.4613012 > [5,] -3.459801 -0.38013895 -0.05011048 -1.183303 -0.5066405 0.1602583 > [6,] 2.282431 2.19639308 1.39946079 4.753675 4.8337840 2.7983964 > V13 V14 V15 V16 V17 V18 > [1,] -4.303745 -1.844632 -3.1044275 -3.4759833 -0.1924864 -1.2015133 > [2,] 3.116199 2.942757 3.2217711 2.9451114 2.8716179 3.4992808 > [3,] -1.627192 -1.779755 -1.5114489 -0.8413003 -0.1924864 0.3091839 > [4,] -3.289903 -2.870566 -3.1055409 -3.5218687 -0.3801390 -1.2399416 > [5,] -4.290241 -2.870453 -2.1055097 -3.5218168 -0.3801390 -1.2400014 > [6,] 3.685705 3.460004 0.8445695 2.4490306 2.6279081 0.5358322 > > sessionInfo() > R version 2.15.1 (2012-06-22) > Platform: x86_64-pc-mingw32/x64 (64-bit) > locale: > [1] LC_COLLATE=English_United States.1252 > [2] LC_CTYPE=English_United States.1252 > [3] LC_MONETARY=English_United States.1252 > [4] LC_NUMERIC=C > [5] LC_TIME=English_United States.1252 > attached base packages: > [1] splines stats graphics grDevices utils datasets methods > [8] base > other attached packages: > [1] scales_0.2.1 Rsamtools_1.8.5 Biostrings_2.24.1 > [4] GenomicRanges_1.8.7 IRanges_1.14.4 BiocGenerics_0.2.0 > [7] cqn_1.2.0 quantreg_4.81 SparseM_0.96 > [10] preprocessCore_1.18.0 nor1mix_1.1-3 mclust_3.5 > loaded via a namespace (and not attached): > [1] bitops_1.0-4.1 colorspace_1.1-1 dichromat_1.2-4 > labeling_0.1 > [5] munsell_0.3 plyr_1.7.1 RColorBrewer_1.0-5 > stats4_2.15.1 > [9] stringr_0.6 zlibbioc_1.2.0 > [[alternative HTML version deleted]]