R/runTSCAN.R
runTSCANClusterDEAnalysis.Rd
This function finds all paths that root from a given cluster
useCluster
, and performs tests to identify significant features for
each path, and are not significant and/or changing in the opposite direction
in the other paths. Using a branching cluster (i.e. a node with degree > 2)
may highlight features which are responsible for the branching event. MST has
to be pre-calculated with runTSCAN
.
runTSCANClusterDEAnalysis(
inSCE,
useCluster,
useAssay = "logcounts",
fdrThreshold = 0.05
)
Input SingleCellExperiment object.
The cluster to be regarded as the root, has to existing in
colData(inSCE)$TSCAN_clusters
.
Character. The name of the assay to use. This assay should
contain log normalized counts. Default "logcounts"
.
Only out put DEGs with FDR value smaller than this value.
Default 0.05
.
The input inSCE
with results updated in metadata
.
data("mouseBrainSubsetSCE", package = "singleCellTK")
mouseBrainSubsetSCE <- runTSCAN(inSCE = mouseBrainSubsetSCE,
useReducedDim = "PCA_logcounts")
#> Sat Mar 18 10:31:32 2023 ... Running 'scran SNN clustering' with 'louvain' algorithm
#> Sat Mar 18 10:31:33 2023 ... Identified 2 clusters
#> Sat Mar 18 10:31:33 2023 ... Running TSCAN to estimate pseudotime
#> Sat Mar 18 10:31:33 2023 ... Clusters involved in path index 2 are: 1, 2
#> Sat Mar 18 10:31:33 2023 ... Number of estimated paths is 1
mouseBrainSubsetSCE <- runTSCANClusterDEAnalysis(inSCE = mouseBrainSubsetSCE,
useCluster = 1)
#> Sat Mar 18 10:31:33 2023 ... Finding DEG between TSCAN branches
#> Sat Mar 18 10:31:33 2023 ... Clusters involved in path index 2 are: 1, 2