Different number of fragments after doublets removal in each run

Question

[Bioinformatics_16]

Hi, Currently im using ArchR to analyzed my scATAC-seq data, after doublets identification and removal each run via using same threshold, same data but we got different number of fragments. is there anyone have idea how i can fix this problem.

[Bastien Hervé]

Could you elaborate a bit more on your issue and add your code to your post ?

[Bioinformatics_16]

AddDoubleScores

doubScores <- addDoubletScores(
    input = ArrowFiles,
    k = 10, #Refers to how many cells near a "pseudo-doublet" to count.
    knnMethod = "LSI", #Refers to the embedding to use for nearest neighbor search with doublet projection.
    LSIMethod = 1,
    threads=4,
    LSIParams = list(
        outlierQuantiles = c(0.02, 0.98), 
        filterBias = FALSE,iterations = 3, 
        UMAPParams = list(
        n_neighbors = 40, 
        min_dist = 0.4, 
        metric = "cosine", 
        verbose = FALSE, 
        fast_sgd = TRUE),
        clusterParams = list( #See Seurat::FindClusters
        resolution = c(1.5), 
        sampleCells = 10000, 
        n.start = 10,
        algorithm=2
    ), 
    varFeatures = 200000, 
    dimsToUse = 1:30),
    force = TRUE
) 

after removed these doublets via using

2. Filtering doublet using default setting

Merged.proj2 <- filterDoublets(Merged.proj,filterRatio = 1.2)

Visualize the output

Merged.proj2 numberOfCells(1): 21335

[Bioinformatics_16]

Thank You for your response.