Thank you so much! Here are the steps we took, as well as the resulting coldata for the ads object, as well as the sce object we converted it from.

# STEP 3: Normalizing the data for DE analysis using Seurat 
samples <- list(mab325C, mab320C, mab3PD20C, mab3PD25c1, mab3PD25c2, wt_24, mab320C_outlierless)
samples <- lapply(samples, function(sample) {
  sample <- NormalizeData(sample)
  sample <- FindVariableFeatures(sample, selection.method = "vst", nfeatures = 2000)

  all.genes <- rownames(sample)
  sample <- ScaleData(sample, features = all.genes)

  max_pcs <- min(30, ncol(sample) - 1)
  if (max_pcs < 2) {
    return(sample)
  }

  sample <- RunPCA(sample, features = VariableFeatures(object = sample), npcs = max_pcs)
  suppressWarnings({
    sample <- RunUMAP(sample, dims = 1:min(10, max_pcs))
  })

  return(sample)
})

merged <- merge(
  x = samples[[1]],       
  y = samples[-1]       
)

# STEP 4: PseudoBulking using https://hbctraining.github.io/scRNA-seq/lessons/pseudobulk_DESeq2_scrnaseq.html
merged_joined <- JoinLayers(merged, assay = "RNA") 
sce <- as.SingleCellExperiment(merged_joined)
sce$sample_id <- factor(sce$orig.ident)

pb <- aggregateAcrossCells(
  sce,
  ids = DataFrame(condition = sce$sample_id)
)

pb_counts <- as.matrix(counts(pb))
colnames(pb_counts) <- pb$condition

pb_conditions <- colnames(pb_counts)
pb_id <- pb_conditions

condition_metadata <- data.frame(
  row.names = pb_conditions,
  pb_id = pb_id
)

dds <- DESeqDataSetFromMatrix(
  countData = pb_counts,
  colData = condition_metadata,
  design = ~ pb_id
)

# Step 5: Running DESeq2
dds <- DESeq(dds)