Skip to content

Commit

Permalink
Add prior count to CPM & add FDR order again
Browse files Browse the repository at this point in the history
  • Loading branch information
@tomkuipers1402
tomkuipers1402 committed Feb 26, 2021
1 parent 85a3bfb commit 0216e4d
Show file tree
Hide file tree
Showing 3 changed files with 15 additions and 18 deletions.
15 changes: 5 additions & 10 deletions inst/src/markdown/analysisDESeq2.Rmd
View file
Original file line number Diff line number Diff line change
Expand Up @@ -114,7 +114,7 @@ dge <- DGEList(counts = assay(se),
dge <- dge[rowSums(abs(dge$counts)) > 1,]
tempDge <- dge
tempDge$counts <- cpm(dge, log = TRUE)
tempDge$counts <- cpm(dge, log = TRUE, prior.count = 1)
countDistributionLinePlot(tempDge)
dge <- DGEList(counts = assay(se), samples = colData(se))
Expand All @@ -137,7 +137,7 @@ analysis <- DESeqDataSet(se, design = design)
analysis <- analysis[rowSums(abs(assay(analysis))) > 1,]
# FILTER IF NESSECARY
counts <- cpm(counts(analysis), log = TRUE)
counts <- cpm(counts(analysis), log = TRUE, prior.count = 1)
selectedFeatures <- rownames(analysis)[apply(counts, 1, function(v)
sum(v >= cpm_value)) >= 1 / 4 * ncol(counts)]
analysis <- analysis[selectedFeatures, ]
Expand All @@ -161,10 +161,8 @@ Extract the normalized data from the analysis results.
# GET NORMALIZED COUNTS
getSize <- estimateSizeFactors(analysis)
normCounts <- cpm(data.frame(counts(getSize, normalized = TRUE)))
normDge <- DGEList(counts = normCounts, samples = dge$samples)
normDge$counts <- log2(normDge$counts)
normDge$counts[is.infinite(normDge$counts)] <- 0
normDge <- DGEList(counts = data.frame(counts(getSize, normalized = TRUE)), samples = dge$samples)
normDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
countDistributionLinePlot(normDge)
samplePca2dPlot(normDge, design_base, "PC1", "PC2")
Expand Down Expand Up @@ -217,10 +215,6 @@ if (!is.null(data_annotation)) {
deTab$Row.names <- NULL
}
# MAKE BOTH TABLES EQUAL
deTab <- deTab[intersect(rownames(deTab), rownames(normDge$counts)), ]
normDge$counts <- normDge$counts[intersect(rownames(deTab), rownames(normDge$counts)), ]
#ORDER deTab TABLE
deTab <- rename(deTab, "FDR" = "adj.P.Val")
deOrder <- c("avgLog2CPM", "avgLog2FC", "P.Value", "FDR", "DE")
Expand Down Expand Up @@ -264,6 +258,7 @@ Results are saved, so they can be retrieved by the application.
```{r save}
# SAVE ANALYSIS
deTab <- deTab[order(deTab$FDR),]
save(deTab, normDge, file = "analysis.RData")
```
Expand Down
9 changes: 5 additions & 4 deletions inst/src/markdown/analysisEdgeR.Rmd
View file
Original file line number Diff line number Diff line change
Expand Up @@ -122,7 +122,7 @@ row.names(dge$genes) <- row.names(dge$counts)
dge <- dge[rowSums(abs(dge$counts)) > 1,]
tempDge <- dge
tempDge$counts <- cpm(dge, log = TRUE)
tempDge$counts <- cpm(dge, log = TRUE, prior.count = 1)
countDistributionLinePlot(tempDge)
```
Expand All @@ -134,7 +134,7 @@ The raw data is filtered based on the input values provided.
# GET SELECTED FEATURES
edger <- calcNormFactors(dge, method = "TMM")
counts <- cpm(edger, log = TRUE)
counts <- cpm(edger, log = TRUE, prior.count = 1)
selectedFeatures <- rownames(edger)[apply(counts, 1, function(v)
sum(v >= cpm_value)) >= 1 / 4 * ncol(counts)]
Expand All @@ -157,7 +157,7 @@ The filtered data is normalized using TMM.
normDge <- calcNormFactors(highExprDge, method = "TMM")
tempDge <- normDge
tempDge$counts <- cpm(normDge, log = TRUE)
tempDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
countDistributionLinePlot(tempDge)
samplePca2dPlot(tempDge, design_base, "PC1", "PC2")
Expand Down Expand Up @@ -255,7 +255,8 @@ Results are saved, so they can be retrieved by the application.
```{r save}
# SAVE ANALYSIS
normDge$counts <- cpm(normDge, log = TRUE)
deTab <- deTab[order(deTab$FDR),]
normDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
save(deTab, normDge, file = "analysis.RData")
```
Expand Down
9 changes: 5 additions & 4 deletions inst/src/markdown/analysisLimma.Rmd
View file
Original file line number Diff line number Diff line change
Expand Up @@ -122,7 +122,7 @@ row.names(dge$genes) <- row.names(dge$counts)
dge <- dge[rowSums(abs(dge$counts)) > 1,]
tempDge <- dge
tempDge$counts <- cpm(dge, log = TRUE)
tempDge$counts <- cpm(dge, log = TRUE, prior.count = 1)
countDistributionLinePlot(tempDge)
```
Expand All @@ -134,7 +134,7 @@ The raw data is filtered based on the input values provided.
# GET SELECTED FEATURES
limmaV <- calcNormFactors(dge, method = "TMM")
counts <- cpm(limmaV, log = TRUE)
counts <- cpm(limmaV, log = TRUE, prior.count = 1)
selectedFeatures <- rownames(limmaV)[apply(counts, 1, function(v)
sum(v >= cpm_value)) >= 1 / 4 * ncol(counts)]
Expand All @@ -157,7 +157,7 @@ The filtered data is normalized using TMM.
normDge <- calcNormFactors(highExprDge, method = "TMM")
tempDge <- normDge
tempDge$counts <- cpm(normDge, log = TRUE)
tempDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
countDistributionLinePlot(tempDge)
samplePca2dPlot(tempDge, design_base, "PC1", "PC2")
Expand Down Expand Up @@ -257,7 +257,8 @@ Results are saved, so they can be retrieved by the application.
```{r save}
# SAVE ANALYSIS
normDge$counts <- cpm(normDge, log = TRUE)
deTab <- deTab[order(deTab$FDR),]
normDge$counts <- cpm(normDge, log = TRUE, prior.count = 1)
save(deTab, normDge, file = "analysis.RData")
```
Expand Down

0 comments on commit 0216e4d

Please sign in to comment.