In compbiomed/singleCellTK: Comprehensive and Interactive Analysis of Single Cell RNA-Seq Data

require(singleCellTK)
require(DT)
require(ggplot2)

sce <- params$object
study <- params$study
useReducedDim <- params$useReducedDim
metas <- metadata(sce)$diffExp[[study]]
featureDisplay <- params$featureDisplay
useAssay <- metas$useAssay
cond1 <- metas$groupNames[[1]]
cond2 <- metas$groupNames[[2]]
helpURL <- paste0("https://camplab.net/sctk/v", packageVersion("singleCellTK"), "/reference/runDEAnalysis.html")

Analysis: r study

The differential expression was performed by comparing the feature expression of cells belonging to condition "r cond1" against the cells belonging to condition "r cond2". The relationship between the two selected group is could be indicated via the embedding plot below. The feature expression matrix used was "r useAssay".

In the condition "r cond1", r length(which(metas$select$ix1)) cells were included, while r length(which(metas$select$ix2)) cells were included for condition r cond2.

The method used for performing the differential expression analysis was "r metas$method". For the more information of the method, please refer to the help page.

sce$deClass <- "Not involved"
sce$deClass[metas$select$ix1] <- cond1
sce$deClass[metas$select$ix2] <- cond2
twoColors <- discreteColorPalette(2, palette = "ggplot")
colorValue <- c("grey", twoColors[1], twoColors[2])
names(colorValue) <- c("Not involved", cond1, cond2)
plotSCEDimReduceColData(sce, colorBy = "deClass", reducedDimName = useReducedDim, labelClusters = FALSE) + scale_color_manual(values = colorValue)

Result Table

The following table presents the statistic metrics of the differential expression analysis. The DEGs in the table are pre-filtered by FDR value less than 0.05.

resultTable <- getDEGTopTable(sce, useResult = study, log2fcThreshold = 0, labelBy = featureDisplay)
datatable(resultTable, options = list(pageLength = 10, scollX = 400))

The table above can be reordered by clicking on the column name.

Visualization

Heatmap

The following heatmap will be divided by the conditions and the regulation. A gene will be determined as up-regulated in condition r cond1 if its log2FC value was greater than 0.25 and FDR is less than 0.05, and down-regulated in condition r cond1 if the log2FC value is less than -0.25. The gene regulation and condition setting will be annotated by default. In addition, if the condition setting was achieved by using a categorical annotation in the object, the annotation used will also be labeled to the column level of the heatmap.

plotDEGHeatmap(sce, study, rowLabel = featureDisplay)

Volcano Plot

The following volcano plot contains dots as all the genes, while the log2FC value on the X axis and -log10(FDR) value on the Y axis. That's to say, the genes farther away from the center of the plot (i.e. at the left or right) show larger fold-change between the two conditions, while the genes closer to the top of the plot show higher significance. The genes with positive log2FC values are up-regulated in the group r cond1, and those with negative log2FC values are up-regulated in r cond2. 10 genes with top fold-change (absolute value) are labeled with text.

plotDEGVolcano(sce, study, featureDisplay = featureDisplay)

Regression Plot

The following plot shows the linear model regression of the top 36 DEGs, ranked by FDR values.

plotDEGRegression(sce, study, labelBy = featureDisplay)

Violin Plot

The following plot shows the violin plot also for comparing the expression of the top 36 DEGs between the two conditions.

plotDEGViolin(sce, study, labelBy = featureDisplay)

SessionInfo

sessionInfo()

compbiomed/singleCellTK documentation built on May 8, 2024, 6:58 p.m.