Demo Bioconductor 組員 : 江建霖 謝紹陽 李祥豪

 eset <- rma(mydata) 運用 RMA 方法創造標 準化和背景校正符號值。產生的資料被存在 ExpressionSet class in the 'eset' object.For large data sets use the more memory efficient justRMA() function.  eset <- mas5(mydata) 使用 expresso (MAS 5.0 方法 )module 代替 RMA(Robust Multiarray Average) 方法,which is much slower than RMA!

 eset_gcrma <- gcrma(mydata) Use this command to aquire gcrma data.The 'library(gcrma)' needs to be loaded first.  eset_plier <- justPlier(mydata) Use this command to aquire plier data. The 'library(plier)' needs to be loaded first.  eset_PMA <- mas5calls(mydata) Generates MAS 5.0 P/M/A calls. The command creates ExpressionSet with P/M/A calls in the 'exprs' slot and the wilcoxon p- values in the 'se.exprs' slot.

 eset <- expresso(mydata, normalize.method ="invariantset", bg.correct=FALSE, pmcorrect.method="pmonly", summary.method="liwong") Generates expression calls similar to dChip (MBEI) method from Li and Wong  library(affycoretools); affystart(plot=T, express="mas5") Handy function 去標準化所有的 CEL 檔案 in current working directory, 執行 qc plot, and export normalized data to file. Works for mas5, rma and gcrma.  library(affyQCReport); QCReport(mydata, file="ExampleQC.pdf") 產生廣泛的 QC report for the AffyBatch object 'mydata' in PDF format.

 image(mydata[,1]) Reconstructs image with log intensities of first chip.  hist(mydata[,1:2]) Plots histogram of PM (Perfect Match) intensities for 1st and 2nd array.  hist(log2(pm(mydata[,1])), breaks=100, col="blue") Plots bar histogram of the PM ('pm') or MM ('mm') log intensities of 1st array.  boxplot(mydata,col="red") 產生 box plot of un-normalized log intensity values.

 boxplot(data.frame(exprs(eset)),col="b lue", main="Normalized Data") 產生 box plot of normalized log intensity values.  mva.pairs(pm(mydata)[,c(1,4)]) Creates MA-plot for un-normalized data. A MA-plot is a plot of log-intensity ratios (M- values) versus log-intensity averages (A- values) between selected chips (here '[1,4]').  mva.pairs(exprs(eset)[,c(1,4)]) Creates MA-plot for normalized data.

Thanks for your listening