Pre-process the data John Blischak Instructor DataCamp - PowerPoint PPT Presentation

DataCamp Differential Expression Analysis with limma in R DIFFERENTIAL EXPRESSION ANALYSIS WITH LIMMA IN R Pre-process the data John Blischak Instructor

DataCamp Differential Expression Analysis with limma in R Mechanism of doxorubicin-induced cardiotoxicity 2x2 design to study mechanism of doxorubicin: 2 genotypes: wild type ( wt ), Top2b null ( top2b ) 2 treatments: PBS ( pbs ), doxorubicin ( dox ) Zhang et al. 2012 dim(eset) Features Samples 29532 12 table(pData(eset)[, c("genotype", "treatment")]) treatment genotype dox pbs top2b 3 3 wt 3 3

DataCamp Differential Expression Analysis with limma in R Inspect the features plotDensities(eset, group = pData(eset)[, "genotype"], legend = "topright")

DataCamp Differential Expression Analysis with limma in R Pre-processing steps Log transform Quantile normalize Filter

DataCamp Differential Expression Analysis with limma in R Sanity check: Boxplot of Top2b boxplot(<y-axis> ~ <x-axis>, main = "<title>") boxplot(<gene expression> ~ <phenotype>, main = "<feature>") boxplot(<Top2b expression> ~ <genotype>, main = "<Top2b info>")

DataCamp Differential Expression Analysis with limma in R Check sources of variation Principal components analysis limma function plotMDS How many clusters do you anticipate?

DataCamp Differential Expression Analysis with limma in R DIFFERENTIAL EXPRESSION ANALYSIS WITH LIMMA IN R Let's practice!

DataCamp Differential Expression Analysis with limma in R DIFFERENTIAL EXPRESSION ANALYSIS WITH LIMMA IN R Model the data John Blischak Instructor

DataCamp Differential Expression Analysis with limma in R Ready for analysis # Plot principal components labeled by genotype plotMDS(eset, labels = pData(eset)[, "genotype"], gene.selection = "common") # Plot principal components labeled by treatment plotMDS(eset, labels = pData(eset)[, "treatment"], gene.selection = "common")

DataCamp Differential Expression Analysis with limma in R Steps for differential expression analysis Build the design matrix with model.matrix Contruct the contrasts matrix with makeContrasts Test the contrasts with lmFit , contrasts.fit , and eBayes

DataCamp Differential Expression Analysis with limma in R Group-means model for doxorubicin study Y = β X + β X + β X + β X + ϵ 1 1 2 2 3 3 4 4 β - Mean expression level in top2b mice treated with dox 1 β - Mean expression level in top2b treated with pbs 2 β - Mean expression level in wt mice treated with dox 3 β - Mean expression level in wt mice treated with pbs 4

DataCamp Differential Expression Analysis with limma in R Contrasts for doxorubicin study β β β β 1 2 3 4 top2b top2b wt wt genotype dox pbs dox pbs treatment Response of wild type mice to dox treatment: β − β = 0 3 4 Response of Top2b null mice to dox treatment: β − β = 0 1 2 Differences between Top2b null and wild type mice in response to dox treatment: ( β − β ) − ( β − β ) = 0 1 2 3 4

DataCamp Differential Expression Analysis with limma in R Testing the doxorubicin study Fit the model coefficients with lmFit Fit the contrasts with contrasts.fit Calculate the t-statistics with eBayes # Summarize results results <- decideTests(fit2) summary(results) # Create a Venn diagram vennDiagram(results)

DataCamp Differential Expression Analysis with limma in R DIFFERENTIAL EXPRESSION ANALYSIS WITH LIMMA IN R Let's practice!

DataCamp Differential Expression Analysis with limma in R DIFFERENTIAL EXPRESSION ANALYSIS WITH LIMMA IN R Inspect the results John Blischak Instructor

DataCamp Differential Expression Analysis with limma in R Inspect the results # Create a Venn diagram vennDiagram(results)

DataCamp Differential Expression Analysis with limma in R Histograms of p-values hist(runif(10000)) limma function topTable Specify contrast coef = "dox_wt" coef = "dox_top2b" coef = "interaction"

DataCamp Differential Expression Analysis with limma in R Volcano plot limma function volcanoplot x-axis: log-fold change in expression between contrasted groups y-axis: log odds of differential expression Specify contrast, e.g. coef = "dox_wt"

DataCamp Differential Expression Analysis with limma in R Testing for KEGG enrichment limma function kegga and topKEGG Specify contrast, e.g. coef = "dox_wt" Specify mouse species with species = "Mm" Mm == Mus musculus

DataCamp Differential Expression Analysis with limma in R DIFFERENTIAL EXPRESSION ANALYSIS WITH LIMMA IN R Let's practice!

DataCamp Differential Expression Analysis with limma in R DIFFERENTIAL EXPRESSION ANALYSIS WITH LIMMA IN R Conclusion John Blischak Instructor

DataCamp Differential Expression Analysis with limma in R Pre-processing features Visualize with plotDensities 3 steps: Log transform Normalize Filter

DataCamp Differential Expression Analysis with limma in R Visualize single genes with boxplot top2b <- which(fData(eset)[, "symbol"] == "Top2b") boxplot(exprs(eset)[top2b, ] ~ pData(eset)[, "genotype"], main = fData(eset)[top2b, ])

DataCamp Differential Expression Analysis with limma in R Check sources of variation # Plot principal components labeled by genotype plotMDS(eset, labels = pData(eset)[,"genotype"], gene.selection = "common") # Plot principal components labeled by treatment plotMDS(eset, labels = pData(eset)[,"treatment"], gene.selection = "common")

DataCamp Differential Expression Analysis with limma in R Flexibly testing various study designs # Create single variable group <- with(pData(eset), paste(genotype, treatment, sep = ".")) group <- factor(group) # Create design matrix with no intercept design <- model.matrix(~0 + group) colnames(design) <- levels(group) # Create a contrasts matrix cm <- makeContrasts(dox_wt = wt.dox - wt.pbs, dox_top2b = top2b.dox - top2b.pbs, interaction = (top2b.dox - top2b.pbs) - (wt.dox - wt.pbs), levels = design) # Fit the model fit <- lmFit(eset, design) # Fit the contrasts fit2 <- contrasts.fit(fit, contrasts = cm) # Calculate the t-statistics for the contrasts fit2 <- eBayes(fit2)

DataCamp Differential Expression Analysis with limma in R Histograms of p-values topTable and hist

DataCamp Differential Expression Analysis with limma in R Volcano plots # Extract the gene symbols gene_symbols <- fit2$genes[, "symbol"] # Create a volcano plot for the contrast dox_wt volcanoplot(fit2, coef = "dox_wt", highlight = 5, names = gene_symbols)

DataCamp Differential Expression Analysis with limma in R Test for enrichment of gene sets # Extract the entrez gene IDs entrez <- fit2$genes[, "entrez"] # Test for enriched KEGG Pathways for contrast dox_wt enrich_dox_wt <- kegga(fit2, coef = "dox_wt", geneid = entrez, species = "Mm") # View the top 5 enriched KEGG pathways topKEGG(enrich_dox_wt, number = 5) Pathway path:mmu05322 Systemic lupus erythematosus path:mmu03008 Ribosome biogenesis in eukaryotes path:mmu05034 Alcoholism path:mmu05412 Arrhythmogenic right ventricular cardiomyopathy (ARVC) path:mmu05330 Allograft rejection N Up Down P.Up P.Down path:mmu05322 76 37 1 3.657708e-10 9.999999e-01 path:mmu03008 71 34 4 3.320811e-09 9.997410e-01 path:mmu05034 130 47 17 2.358456e-07 9.733029e-01 path:mmu05412 52 2 26 9.995025e-01 4.140466e-07 path:mmu05330 26 16 0 6.720834e-07 1.000000e+00

DataCamp Differential Expression Analysis with limma in R DIFFERENTIAL EXPRESSION ANALYSIS WITH LIMMA IN R Congrats on completing the course!