No peaks found in FindMarker output

[ahmadsam66]

Hello,

I have worked on a 10x sample with ~4000 cells. But I faced with a problem. With different resolution I can't see any significant peaks in FindMarkers functions. For example res 0.3 = the first big cluster has no peaks..... and e.g res. 0.8 from 14 clusters 4 of them has no significant peaks output, although in RNA level I could found markers in all resolutions.

What is the problem with this analysis ?
How can I fix it?

my code is attached below :

#readRDS("~/Documents/Projects/ATAC-Seq/sc_atac.rds")
newpath="~/Documents/Projects/ATAC-Seq/"
library(rGREAT)
library(ggplot2)
library(Signac)
library(Seurat)
library(GenomeInfoDb)
library(EnsDb.Mmusculus.v75)
library(ggplot2)
library(patchwork)
set.seed(1234)

counts <- Read10X_h5(filename = "~/Documents/filtered_peak_bc_matrix.h5")

metadata <- read.csv(
  file="~/Documents/singlecell.csv",
  header = TRUE,
  row.names = 1
)

chrom_assay <- CreateChromatinAssay(
  counts = counts,
  sep = c(":", "-"),
  genome = 'mm10',
  fragments = '~/Documents/fragments.tsv.gz',
  min.cells = 10,
  min.features = 200
)

pbmc <- CreateSeuratObject(
  counts = chrom_assay,
  assay = "peaks",  meta.data = metadata 
)

annotations <- GetGRangesFromEnsDb(ensdb =EnsDb.Mmusculus.v75)

# change to UCSC style since the data was mapped to hg19
seqlevelsStyle(annotations) <- 'UCSC'
genome(annotations) <- "mm10"

# add the gene information to the object
Annotation(pbmc) <- annotations

pbmc <- NucleosomeSignal(object = pbmc)

# compute TSS enrichment score per cell
pbmc <- TSSEnrichment(object = pbmc, fast = FALSE)

# add blacklist ratio and fraction of reads in peaks
pbmc$pct_reads_in_peaks <- pbmc$peak_region_fragments / pbmc$passed_filters * 100
pbmc$blacklist_ratio <- pbmc$blacklist_region_fragments / pbmc$peak_region_fragments


pbmc$high.tss <- ifelse(pbmc$TSS.enrichment > 2, 'High', 'Low')
TSSPlot(pbmc, group.by = 'high.tss') + NoLegend()


makeViolin <- function (x)
{
VlnPlot(
  object = pbmc,
  features =x ,
  pt.size = 0.1
)
}

metrics <- c('pct_reads_in_peaks', 'peak_region_fragments',
               'TSS.enrichment', 'blacklist_ratio', 'nucleosome_signal')
lapply(metrics, makeViolin)



# filtering data

pbmc <- subset(
  x = pbmc,
  subset = peak_region_fragments > 12000 &
    peak_region_fragments < 65000 &
    pct_reads_in_peaks > 15 &
    blacklist_ratio < 0.05 &
    nucleosome_signal < 4 &
    TSS.enrichment > 2
)
pbmc


tumor <- pbmc
pbmc <- RunTFIDF(pbmc)
pbmc <- FindTopFeatures(pbmc, min.cutoff = 'q0')
pbmc <- RunSVD(pbmc)
DepthCor(pbmc)


# I used resolution 0.6 for clustering.

pbmc <- RunUMAP(object = pbmc, reduction = 'lsi', dims = 2:30)
pbmc <- FindNeighbors(object = pbmc, reduction = 'lsi', dims = 2:30)
pbmc <- FindClusters(object = pbmc, verbose = FALSE, algorithm = 3,resolution = 0.6)
#DimPlot(object = pbmc, label = TRUE) + NoLegend()
DimPlot(object = pbmc, label = TRUE,label.size = 8) 


gene.activities <- GeneActivity(pbmc)

DefaultAssay(pbmc) <- 'peaks'
da_peaks<- FindAllMarkers( object = pbmc, min.pct = 0.1, test.use = 'LR', latent.vars = 'peak_region_fragments')

The point is the da_peaks for cluster 0 has no significant peaks.
Thanks.

[ahmadsam66]

No suggestions !?

[timoast]

Hi @ahmadsam66

It's difficult to guess what the issue could be here without any code and more information about your experiment. One potential issue is that clusters could be poorly-defined, with few genuine differentially accessible peaks between them

[ahmadsam66]

I changed the post. I added my code to it.

[timoast]

You could try contrasting specific clusters rather than using FindAllMarkers(), which compares one cluster to all others

[ahmadsam66]

I also try it individually but the result was same, when I ran FinadMarkers() for cluster 0 it gave zero peak.