Thank you to the author for the quick reply. If I use gene_counts_length_scaled.tsv as a metric that has already normalized for library size and the average transcript length, what additional components will dds <- DESeq(dds) continue to normalize? Will this affect my final results? I think that pre-normalized counts (like length-scaled counts) can lead to over-normalization when DESeq2 applies its own methods on top of existing adjustments, potentially distorting expression values.

I want to observe which genes are upregulated and downregulated, so I need the log2FoldChange and padj. Below is the nf-core code.

# Define pattern for file names based on quantification type
pattern <- ifelse('$quant_type' == "kallisto", "abundance.tsv", "quant.sf")
fns <- list.files('quants', pattern = pattern, recursive = T, full.names = T)
names <- basename(dirname(fns))
names(fns) <- names
dropInfReps <- '$quant_type' == "kallisto"

# Import transcript-level quantifications
txi <- tximport(fns, type = '$quant_type', txOut = TRUE, dropInfReps = dropInfReps)

# Read transcript and sample data
transcript_info <- read_transcript_info('$tx2gene')

# Make coldata just to appease the summarizedexperiment
coldata <- data.frame(files = fns, names = names)
rownames(coldata) <- coldata[["names"]]

# Create initial SummarizedExperiment object
se <- create_summarized_experiment(txi[["counts"]], txi[["abundance"]], txi[["length"]],
    DataFrame(coldata), transcript_info\$transcript)

# Setting parameters for writing tables
params <- list(
    list(obj = se, slot = "abundance", suffix = "transcript_tpm.tsv"),
    list(obj = se, slot = "counts", suffix = "transcript_counts.tsv"),
    list(obj = se, slot = "length", suffix = "transcript_lengths.tsv")
)

# Process gene-level data if tx2gene mapping is available
if ("tx2gene" %in% names(transcript_info) && !is.null(transcript_info\$tx2gene)) {
    tx2gene <- transcript_info\$tx2gene
    gi <- summarizeToGene(txi, tx2gene = tx2gene)
    gi.ls <- summarizeToGene(txi, tx2gene = tx2gene, countsFromAbundance = "lengthScaledTPM")
    gi.s <- summarizeToGene(txi, tx2gene = tx2gene, countsFromAbundance = "scaledTPM")

    gene_info <- transcript_info\$gene[match(rownames(gi[[1]]), transcript_info\$gene[["gene_id"]]),]
    rownames(gene_info) <- NULL
    col_data_frame <- DataFrame(coldata)

    # Create gene-level SummarizedExperiment objects
    gse <- create_summarized_experiment(gi[["counts"]], gi[["abundance"]], gi[["length"]],
        col_data_frame, gene_info)
    gse.ls <- create_summarized_experiment(gi.ls[["counts"]], gi.ls[["abundance"]], gi.ls[["length"]],
        col_data_frame, gene_info)
    gse.s <- create_summarized_experiment(gi.s[["counts"]], gi.s[["abundance"]], gi.s[["length"]],
        col_data_frame, gene_info)

    params <- c(params, list(
        list(obj = gse, slot = "length", suffix = "gene_lengths.tsv"),
        list(obj = gse, slot = "abundance", suffix = "gene_tpm.tsv"),
        list(obj = gse, slot = "counts", suffix = "gene_counts.tsv"),
        list(obj = gse.ls, slot = "counts", suffix = "gene_counts_length_scaled.tsv"),
        list(obj = gse.s, slot = "counts", suffix = "gene_counts_scaled.tsv")
    ))
}