Hi everyone, I'm currently attempting to adapt the SWAN normalization scripts found in the Minfi package for use with MethyLumiM. I noticed this line in the "preprocessSWAN" script and I'm not sure what this is doing. bg <- bgIntensitySwan(rgSet) Here's the bgIntesitySWAN function: bgIntensitySwan <- function (rgSet) { grnMed <- colMedians(getGreen(rgSet)[getControlAddress(rgSet, controlType = "NEGATIVE"), ]) redMed <- colMedians(getRed(rgSet)[getControlAddress(rgSet, controlType = "NEGATIVE"), ]) return(rowMeans(cbind(grnMed, redMed))) } It seems to be calculating the median NEGATIVE signals of each sample in each color channel separately then calculating the average per sample. How is this information used in the quantile normalization? I initially thought the entire SWAN process would only depend on the sample itself and will not be affected by other samples since it's correcting the discordance between Type I and II probes within each sample. Thanks in advance for any insights into this! Cheers, Lucia *Lucia Lam, B Sc * Lab Manager and Research Assistant for Dr. Michael S. Kobor *Centre for Molecular Medicine and Therapeutics (CMMT) ****- Researching Life to Change Lives* *University of British Columbia*** * *> sessionInfo() R version 2.15.0 (2012-03-30) 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] stats graphics grDevices utils datasets methods base other attached packages: [1] IlluminaHumanMethylation450k.db_1.4.6 org.Hs.eg.db_2.7.1 [3] RSQLite_0.11.1 DBI_0.2-5 [5] AnnotationDbi_1.18.1 Biostrings_2.24.1 [7] minfi_1.2.0 GenomicRanges_1.8.6 [9] IRanges_1.14.3 reshape_0.8.4 [11] plyr_1.7.1 lattice_0.20-6 [13] lumi_2.8.0 nleqslv_1.9.3 [15] methylumi_2.2.0 ggplot2_0.9.1 [17] reshape2_1.2.1 scales_0.2.1 [19] Biobase_2.16.0 BiocGenerics_0.2.0 loaded via a namespace (and not attached): [1] affy_1.34.0 affyio_1.24.0 annotate_1.34.0 [4] beanplot_1.1 bigmemory_4.2.11 BiocInstaller_1.4.7 [7] bit_1.1-8 bitops_1.0-4.1 BSgenome_1.24.0 [10] codetools_0.2-8 colorspace_1.1-1 crlmm_1.14.3 [13] dichromat_1.2-4 digest_0.5.2 DNAcopy_1.30.0 [16] ellipse_0.3-7 ff_2.2-7 foreach_1.4.0 [19] genefilter_1.38.0 genoset_1.6.0 grid_2.15.0 [22] hdrcde_2.16 iterators_1.0.6 KernSmooth_2.23-7 [25] labeling_0.1 limma_3.12.0 MASS_7.3-18 [28] Matrix_1.0-6 matrixStats_0.5.0 mclust_3.4.11 [31] memoise_0.1 mgcv_1.7-17 multtest_2.12.0 [34] munsell_0.3 mvtnorm_0.9-9992 nlme_3.1-104 [37] nor1mix_1.1-3 oligoClasses_1.18.0 preprocessCore_1.18.0 [40] proto_0.3-9.2 R.methodsS3_1.4.2 RColorBrewer_1.0-5 [43] RCurl_1.91-1.1 Rsamtools_1.8.5 rtracklayer_1.16.1 [46] siggenes_1.30.0 splines_2.15.0 stats4_2.15.0 [49] stringr_0.6 survival_2.36-14 tools_2.15.0 [52] XML_3.9-4.1 xtable_1.7-0 zlibbioc_1.2.0 [[alternative HTML version deleted]]

Sorry for the redundant post: Just revisited the scripts again and got some new insights. Please correct me if I'm wrong, but seems like during the subset quantile normalization steps, if the intensities were less than zero after normalization, they were forced to take up the mean negative signals as calculated by bgIntensitySwan? Thanks, Lucia On Tue, Jul 3, 2012 at 4:33 PM, Lucia Lam <lucia@cmmt.ubc.ca> wrote: > Hi everyone, > > I'm currently attempting to adapt the SWAN normalization scripts found in > the Minfi package for use with MethyLumiM. I noticed this line in the > "preprocessSWAN" script and I'm not sure what this is doing. > bg <- bgIntensitySwan(rgSet) > > Here's the bgIntesitySWAN function: > bgIntensitySwan <- function (rgSet) > { > grnMed <- colMedians(getGreen(rgSet)[getControlAddress(rgSet, > controlType = "NEGATIVE"), ]) > redMed <- colMedians(getRed(rgSet)[getControlAddress(rgSet, > controlType = "NEGATIVE"), ]) > return(rowMeans(cbind(grnMed, redMed))) > } > > It seems to be calculating the median NEGATIVE signals of each sample in > each color channel separately then calculating the average per sample. How > is this information used in the quantile normalization? I initially thought > the entire SWAN process would only depend on the sample itself and will not > be affected by other samples since it's correcting the discordance between > Type I and II probes within each sample. > > Thanks in advance for any insights into this! > Cheers, > Lucia > > *Lucia Lam, B Sc > * > > Lab Manager and Research Assistant for Dr. Michael S. Kobor > > *Centre for Molecular Medicine and Therapeutics (CMMT) ****- Researching > Life to Change Lives* > > *University of British Columbia*** > * > > *> sessionInfo() > R version 2.15.0 (2012-03-30) > 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] stats graphics grDevices utils datasets methods base > > other attached packages: > [1] IlluminaHumanMethylation450k.db_1.4.6 > org.Hs.eg.db_2.7.1 > [3] RSQLite_0.11.1 > DBI_0.2-5 > [5] AnnotationDbi_1.18.1 > Biostrings_2.24.1 > [7] minfi_1.2.0 > GenomicRanges_1.8.6 > [9] IRanges_1.14.3 > reshape_0.8.4 > [11] plyr_1.7.1 > lattice_0.20-6 > [13] lumi_2.8.0 > nleqslv_1.9.3 > [15] methylumi_2.2.0 > ggplot2_0.9.1 > [17] reshape2_1.2.1 > scales_0.2.1 > [19] Biobase_2.16.0 > BiocGenerics_0.2.0 > > loaded via a namespace (and not attached): > [1] affy_1.34.0 affyio_1.24.0 annotate_1.34.0 > [4] beanplot_1.1 bigmemory_4.2.11 BiocInstaller_1.4.7 > [7] bit_1.1-8 bitops_1.0-4.1 BSgenome_1.24.0 > [10] codetools_0.2-8 colorspace_1.1-1 crlmm_1.14.3 > [13] dichromat_1.2-4 digest_0.5.2 DNAcopy_1.30.0 > [16] ellipse_0.3-7 ff_2.2-7 foreach_1.4.0 > [19] genefilter_1.38.0 genoset_1.6.0 grid_2.15.0 > [22] hdrcde_2.16 iterators_1.0.6 KernSmooth_2.23-7 > [25] labeling_0.1 limma_3.12.0 MASS_7.3-18 > [28] Matrix_1.0-6 matrixStats_0.5.0 mclust_3.4.11 > [31] memoise_0.1 mgcv_1.7-17 multtest_2.12.0 > [34] munsell_0.3 mvtnorm_0.9-9992 nlme_3.1-104 > [37] nor1mix_1.1-3 oligoClasses_1.18.0 preprocessCore_1.18.0 > [40] proto_0.3-9.2 R.methodsS3_1.4.2 RColorBrewer_1.0-5 > [43] RCurl_1.91-1.1 Rsamtools_1.8.5 rtracklayer_1.16.1 > [46] siggenes_1.30.0 splines_2.15.0 stats4_2.15.0 > [49] stringr_0.6 survival_2.36-14 tools_2.15.0 > [52] XML_3.9-4.1 xtable_1.7-0 zlibbioc_1.2.0 > [[alternative HTML version deleted]]