Possible batch effect in ATAC clusters in scMultiome data

[kryptonitecircuit2002]

I am analysing an scMultiome data using the following code:

#Read files
input.10X.Control <- Read10X_h5("D:/Transit/Single_Cell+ATAC/Combined_Analysis/Control_sample/filtered_feature_bc_matrix_control.h5")
fragpath.Control <- "D:/Transit/Single_Cell+ATAC/Combined_Analysis/Control_sample/atac_fragments_control.tsv.gz"
metadata.Control <- read.csv(
  file = "D:/Transit/Single_Cell+ATAC/Combined_Analysis/Control_sample/per_barcode_metrics_control.csv",
  header = TRUE,
  row.names = 1
)
## Create Seurat object
Control <- CreateSeuratObject(counts = input.10X.Control$`Gene Expression`, 
                           assay = 'RNA',
                           project = sampleID,
                           meta.data = metadata.Control)

Control[["percent.mt"]] <- PercentageFeatureSet(Control, pattern = "^MT-")

#Get annotations 
#annotaion from GTF
gref.path = "D:/Transit/Single_Cell+ATAC/Control/ATAC/Data+Analysis/Analysis.Control/Danio_rerio.GRCz11.105.filtered.gtf"
gene.coords.zf <- rtracklayer::import(gref.path)
genome(gene.coords.zf) <- 'GRCz11'
gene.coords.zf <- gene.coords.zf[! is.na(gene.coords.zf$gene_name),]

##Add ATAC data
atac_counts <- input.10X.Control$Peaks
grange.counts <- StringToGRanges(rownames(atac_counts), sep = c(":", "-"))
grange.use <- seqnames(grange.counts) %in% standardChromosomes(grange.counts)
atac_counts <- atac_counts[as.vector(grange.use), ]
Control[['ATAC']] <- CreateChromatinAssay(
  counts = atac_counts,
  sep = c(":", "-"),
  genome = 'GRCz11',
  fragments = fragpath.Control,
  min.cells = 10,
  annotation = gene.coords.zf
)
#QC
DefaultAssay(Control) <- "ATAC"
Control <- NucleosomeSignal(Control)
Control <- TSSEnrichment(Control, fast = FALSE)
Control$pct_reads_in_peaks <- Control$atac_peak_region_fragments / Control$atac_fragments * 100

#Peak calling
system('bash -c "/home/ayush/.local/bin/macs2 callpeak -t /mnt/d/Transit/Single_Cell+ATAC/Combined_Analysis/Control_sample/atac_fragments_control.tsv.gz -g 1.4e+09 -f BED --nomodel --extsize 200 --shift -100 -n Control_cerbx_script --outdir /mnt/d/Transit/Single_Cell+ATAC/Control/ATAC/Data+Analysis/Analysis.Control/Cerebrocortex" ', 
                wait = TRUE,  ignore.stderr = FALSE,  ignore.stdout = FALSE)
peak_path <- "D:/Transit/Single_Cell+ATAC/Control/ATAC/Data+Analysis/Analysis.Control/Cerebrocortex/Control_cerbx_script_peaks.narrowPeak"
peaks <- rtracklayer::import(peak_path, format = "narrowPeak")
peaks <- keepStandardChromosomes(peaks, pruning.mode = 'coarse')

# quantify counts in each peak
macs_count <- FeatureMatrix(fragments = Fragments(Control),
                            features = peaks,
                            cells = colnames(Control))

Control[['ATAC']] <- CreateChromatinAssay(
  counts = macs_count,
  sep = c(":", "-"),
  genome = 'GRCz11',
  fragments = fragpath.Control,
  min.cells = 10,
  annotation = gene.coords.zf
)

Control <- subset(
  x = Control,
  subset = nCount_ATAC < 50000 &
    nCount_RNA < 10000 &
    nucleosome_signal < 1.5 &
    TSS.enrichment > 1 &
    TSS.enrichment < 10)

#preprocess <- function(Control, n.pc = 30, n.lsi = 10){
  
 #Filter peaks/genes 
  
 tmp <- Matrix::rowSums(Control[["RNA"]]@layers$counts > 0)
 Control[["RNA"]] <- subset(Control[["RNA"]], features = names(which(tmp >= 10)))
 tmp <- Matrix::rowSums(Control[["ATAC"]]@counts > 0)
 Control[["ATAC"]] <- subset(Control[["ATAC"]], features = names(which(tmp >= 10)))
  
# ATAC-seq
  DefaultAssay(Control) <- "ATAC"
  Control <- RunTFIDF(Control, method = 3)
  Control <- FindTopFeatures(Control, min.cutoff = 'q75')
  Control <- RunSVD(Control)
  Control <- RunUMAP(Control, reduction = 'lsi', dims = 2:10, assay = 'ATAC',
                    reduction.name = "umap.atac", reduction.key = "atacUMAP_")
  Control <- FindNeighbors(Control, reduction = 'lsi', dims = 2:10, assay = 'ATAC')
  Control <- FindClusters(Control, graph.name = 'ATAC_snn', algorithm = 3, resolution = 0.3)

However, when I am running the code:
DimPlot(Control, reduction = 'umap.atac', label = T)

I am getting the following ATAC clusters:

Is this arising because of possible batch effects? How do I solve this?