Instructor Notes
FIXME
RNA-seq入門RNA-seq実験では何を測定するのか？実験デザインの考察RNA-seq定量化：リードからカウントマトリックスまで参照配列の検索このワークショップで我々はどこに向かっているのか？
RStudioプロジェクトと実験データ
Rに量的データをインポートして注釈を付ける
Exploratory analysis and quality control
Differential expression analysis
Instructor Note
The function to generate a DESeqDataSet needs to be
adapted depending on the input type, e.g,
R
#From SummarizedExperiment object
ddsSE <- DESeqDataSet(se, design = ~ sex + time)
#From count matrix
dds <- DESeqDataSetFromMatrix(countData = assays(se)$counts,
colData = colData(se),
design = ~ sex + time)
Instructor Note
DESeq2 uses the “Relative Log Expression” (RLE) method to calculate sample-wise size factors tĥat account for read depth and library composition. edgeR uses the “Trimmed Mean of M-Values” (TMM) method to account for library size differences and compositional effects. edgeR’s normalization factors and DESeq2’s size factors yield similar results, but are not equivalent theoretical parameters.
Instructor Note
Both of the below ways of specifying the contrast are essentially
equivalent. The name parameter can be accessed using
resultsNames().
R
resTime <- results(dds, contrast = c("time", "Day8", "Day0"))
resTime <- results(dds, name = "time_Day8_vs_Day0")