Hi All, I am using crlmm package to load Affymetrix SNP6.0 genotype data. crlmmResult[['SNR']] can be used as the quality of arrays and the value less than 5 is of low quality and should be removed from further analysis. I am wondering how to use the confs(crlmmResult), or, what is the threshold of confs for dropping out markers? Thanks! -- output of sessionInfo(): R version 2.15.2 (2012-10-26) Platform: x86_64-unknown-linux-gnu (64-bit) locale: [1] C attached base packages: [1] stats graphics grDevices utils datasets methods base other attached packages: [1] genomewidesnp6Crlmm_1.0.7 crlmm_1.16.9 [3] preprocessCore_1.20.0 oligoClasses_1.20.0 [5] BiocGenerics_0.4.0 loaded via a namespace (and not attached): [1] AnnotationDbi_1.20.5 Biobase_2.18.0 BiocInstaller_1.8.3 [4] Biostrings_2.26.2 DBI_0.2-5 GenomicRanges_1.10.6 [7] IRanges_1.16.4 Matrix_1.0-9 RSQLite_0.11.2 [10] RcppEigen_0.3.1.2.1 XML_3.95-0.1 affyio_1.26.0 [13] annotate_1.36.0 bit_1.1-9 codetools_0.2-8 [16] ellipse_0.3-7 ff_2.2-10 foreach_1.4.0 [19] genefilter_1.40.0 grid_2.15.2 iterators_1.0.6 [22] lattice_0.20-10 mvtnorm_0.9-9994 parallel_2.15.2 [25] splines_2.15.2 stats4_2.15.2 survival_2.36-14 [28] tools_2.15.2 xtable_1.7-1 zlibbioc_1.4.0 -- Sent via the guest posting facility at bioconductor.org.

My recommendation is to define a (quantilic) threshold and filter upon that... For example, if your preference is to remove the 1% "worse" calls for your samples: theCalls = calls(crlmmResult) theConfs = confs(crlmmResult) ## identify calls for which the confidence scores ## are smaller than the 1%-quantile, and ignore them thresholds = apply(theConfs, 2, quantile, 0.01) ignore = sweep(theConfs, 2, thresholds, '-') < 0 theCalls[ignore] = NA benilton 2013/3/4 Shrek [guest] <guest at="" bioconductor.org="">: > > Hi All, > I am using crlmm package to load Affymetrix SNP6.0 genotype data. crlmmResult[['SNR']] can be used as the quality of arrays and the value less than 5 is of low quality and should be removed from further analysis. I am wondering how to use the confs(crlmmResult), or, what is the threshold of confs for dropping out markers? > Thanks! > > > -- output of sessionInfo(): > > R version 2.15.2 (2012-10-26) > Platform: x86_64-unknown-linux-gnu (64-bit) > > locale: > [1] C > > attached base packages: > [1] stats graphics grDevices utils datasets methods base > > other attached packages: > [1] genomewidesnp6Crlmm_1.0.7 crlmm_1.16.9 > [3] preprocessCore_1.20.0 oligoClasses_1.20.0 > [5] BiocGenerics_0.4.0 > > loaded via a namespace (and not attached): > [1] AnnotationDbi_1.20.5 Biobase_2.18.0 BiocInstaller_1.8.3 > [4] Biostrings_2.26.2 DBI_0.2-5 GenomicRanges_1.10.6 > [7] IRanges_1.16.4 Matrix_1.0-9 RSQLite_0.11.2 > [10] RcppEigen_0.3.1.2.1 XML_3.95-0.1 affyio_1.26.0 > [13] annotate_1.36.0 bit_1.1-9 codetools_0.2-8 > [16] ellipse_0.3-7 ff_2.2-10 foreach_1.4.0 > [19] genefilter_1.40.0 grid_2.15.2 iterators_1.0.6 > [22] lattice_0.20-10 mvtnorm_0.9-9994 parallel_2.15.2 > [25] splines_2.15.2 stats4_2.15.2 survival_2.36-14 > [28] tools_2.15.2 xtable_1.7-1 zlibbioc_1.4.0 > > > -- > Sent via the guest posting facility at bioconductor.org. > > _______________________________________________ > Bioconductor mailing list > Bioconductor at r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor