doc/batchCorrection.md

In debsin/dropClust: Single Cell Transcriptome Analysis

Integrative analysis

Loading datasets

Each dataset represents one batch and must be a SingleCellExperiment object. The objects are are merged by passing a list in the next step.

library(dropClust)
load(url("https://raw.githubusercontent.com/LuyiTian/CellBench_data/master/data/sincell_with_class.RData"))

objects = list()

objects[[1]] = sce_sc_10x_qc

objects[[2]] = sce_sc_CELseq2_qc

objects[[3]] = sce_sc_Dropseq_qc

Merge datasets using dropClust

Datasets can be merged in two ways: using a set of DE genes from each batch or, using the union of the sets of highly variable genes from each batch.

#> 
#> Procesing Batch 1 ...
#> 1 bad cells removed.
#> 257 genes filtered out, 16211 genes remaining.
#> Sort Top Genes...
#> Cutoff Genes...
#> Building graph with 901 nodes...Louvain Partition...Done.
#> 702 samples extracted.
#> Find best PCA components...[1]  702 1000
#> 200 genes selected.
#> 702 samples and 200 genes used for clustering.
#> Build Graph with 702 samples...Done.
#> Louvain Partitioning...Done.
#> Find nearest neighbours among sub-samples...Done.
#> Post-hoc Cluster Assignment...Done.
#> Unassigned Cells 0 
#> Number of Predicted Clusters: 7 
#> Computing for DE genes:
#> Cluster 1 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 2 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 3 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 4 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 5 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 6 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 7 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> 
#> Procesing Batch 2 ...
#> 1 bad cells removed.
#> 11931 genes filtered out, 16273 genes remaining.
#> Sort Top Genes...
#> Cutoff Genes...
#> Find best PCA components...[1]  273 1000
#> 200 genes selected.
#> 273 samples and 200 genes used for clustering.
#> Build Graph with 273 samples...Done.
#> Louvain Partitioning...Done.
#> Find nearest neighbours among sub-samples...Done.
#> Post-hoc Cluster Assignment...Done.
#> Unassigned Cells 0 
#> Number of Predicted Clusters: 4 
#> Computing for DE genes:
#> Cluster 1 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 2 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 3 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 4 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> 
#> Procesing Batch 3 ...
#> 1 bad cells removed.
#> 958 genes filtered out, 14169 genes remaining.
#> Sort Top Genes...
#> Cutoff Genes...
#> Find best PCA components...[1]  224 1000
#> 200 genes selected.
#> 224 samples and 200 genes used for clustering.
#> Build Graph with 224 samples...Done.
#> Louvain Partitioning...Done.
#> Find nearest neighbours among sub-samples...Done.
#> Post-hoc Cluster Assignment...Done.
#> Unassigned Cells 0 
#> Number of Predicted Clusters: 5 
#> Computing for DE genes:
#> Cluster 1 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 2 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 3 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 4 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 5 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.

Perform correction and dimension reduction

set.seed(1)
dc.corr <-  Correction(merged_data,  method="default", close_th = 0.1, cells_th = 0.1,
                       components = 10, n_neighbors = 30,  min_dist = 1)
#> Batch correcting...
#> from 177 to 100 genes.
#> Embedding with UMAP...Done

Perform Clustering on integrated dimensions

dc.corr = Cluster(dc.corr,method = "kmeans",centers = 3)
#> Clustering on embedded dimensions...Perfom k-means Clustering...Done.
#> Done.

Visualizing clusters

Compute 2D embeddings for samples followed by post-hoc clustering.

ScatterPlot(dc.corr, title = "Clusters")

## Optional Batch correction Users can use fastmnn method for batch correction. Specific arguments of fastmnn can also be passed through the Correction module.

merged_data.fastmnn<-Merge(all.objects,use.de.genes = FALSE)
#> 
#> Procesing Batch 1 ...
#> 1 bad cells removed.
#> 257 genes filtered out, 16211 genes remaining.
#> Sort Top Genes...
#> Cutoff Genes...
#> 
#> Procesing Batch 2 ...
#> 1 bad cells removed.
#> 11931 genes filtered out, 16273 genes remaining.
#> Sort Top Genes...
#> Cutoff Genes...
#> 
#> Procesing Batch 3 ...
#> 1 bad cells removed.
#> 958 genes filtered out, 14169 genes remaining.
#> Sort Top Genes...
#> Cutoff Genes...

set.seed(1)
mnn.corr <-  Correction(merged_data.fastmnn,  method="fastmnn", d = 10)

mnn.corr = Cluster(mnn.corr,method = "kmeans",centers = 3)
#> Clustering on embedded dimensions...Perfom k-means Clustering...Done.
#> Done.

ScatterPlot(mnn.corr, title = "Clusters")

Marker discovery from the merged dataset

de<-FindMarkers(dc.corr,q_th = 0.001, lfc_th = 1.2,nDE = 10)
#> Computing for DE genes:
#> Cluster 1 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 2 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
#> Cluster 3 :
#>  Computing Wilcoxon p values...Done.
#>  Computing Log fold change values...Done.
de$genes.df
#>       cluster_1         cluster_2         cluster_3        
#>  [1,] "ENSG00000179344" "ENSG00000266891" "ENSG00000249395"
#>  [2,] "ENSG00000121858" "ENSG00000100979" "ENSG00000253706"
#>  [3,] "ENSG00000100234" "ENSG00000100867" "ENSG00000135446"
#>  [4,] "ENSG00000149557" "ENSG00000258949" "ENSG00000132432"
#>  [5,] "ENSG00000214548" "ENSG00000221923" "ENSG00000258484"
#>  [6,] "ENSG00000198502" "ENSG00000129991" "ENSG00000135506"
#>  [7,] "ENSG00000231389" "ENSG00000231290" "ENSG00000147889"
#>  [8,] "ENSG00000196735" "ENSG00000170291" "ENSG00000147604"
#>  [9,] "ENSG00000223865" "ENSG00000164744" "ENSG00000154582"
#> [10,] "ENSG00000151632" "ENSG00000237268" "ENSG00000167641"

debsin/dropClust documentation built on Nov. 4, 2019, 10:22 a.m.