DOTSeq

DOTSeq is an R package for identifying differentially translated open reading frames (ORFs) from ribosome profiling (Ribo-seq) and matched RNA-seq datasets. Unlike traditional gene-level approaches, DOTSeq performs analysis at the ORF level, enabling detection of:

• Differential ORF Usage (DOU) — changes in ORF usage within the same gene across conditions.
• Differential Translation Efficiency (DTE) — changes in ribosome loading relative to RNA level across conditions.

DOTSeq models Ribo-seq and RNA-seq read counts using a beta-binomial generalised linear model (GLM) implemented via glmmTMB. It supports experimental designs with multiple conditions, and uses an interaction term (condition:strategy) to isolate translation-specific effects.

Post hoc contrasts are computed using emmeans, and empirical Bayes shrinkage is applied via ashr.

INSTALLATION

# Create a directory for R packages if not already
package_dir <- file.path(Sys.getenv("HOME"), "R/4.5")
dir.create(package_dir, showWarnings = TRUE, recursive = TRUE)
.libPaths(c(package_dir, .libPaths()))

# Install BiocManager if not already available
if (!require("BiocManager", quietly = TRUE))
  install.packages("BiocManager", lib = package_dir)

# For R 4.6, initialise Bioc devel to install DOTSeq, and required packages 
# with automatic update confirmation
BiocManager::install(version="devel")
BiocManager::install("DOTSeq", lib = package_dir, ask = FALSE)

# For R 4.5
options(repos = BiocManager::repositories(version = "3.22"))
BiocManager::install("compgenom/DOTSeq", lib = package_dir, ask = FALSE)

DOCUMENTATION

TL;DR: To understand how to use DOTSeq without going through the preprocessing steps, please refer to the vignettes.

Preprocessing Steps Required Before Running DOTSeq:

Step 1. Align Ribo-seq and RNA-seq reads

We use a publicly available HeLa cell cycle dataset from Ly 2024.

# Clone DOTSeq repository
git clone https://github.com/compgenom/DOTSeq.git

# Assume FASTQ files are downloaded via SRA Toolkit and stored in:
# DOTSeq/inst/extdata/ly_2024
# STAR index will be generated in:
# DOTSeq/inst/extdata/hg38_star_index

# Download GENCODE annotation and transcript FASTA files
wget https://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_47/gencode.v47.annotation.gtf.gz \
  -O DOTSeq/inst/extdata/gencode.v47.annotation.gtf.gz

wget https://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_47/gencode.v47.pc_transcripts.fa.gz \
  -O DOTSeq/inst/extdata/gencode.v47.pc_transcripts.fa.gz

# Assume reference genome FASTA is downloaded
# Generate STAR index
STAR --runMode genomeGenerate \
  --runThreadN 32 \
  --genomeFastaFiles DOTSeq/inst/extdata/hg38.fa \
  --sjdbGTFfile DOTSeq/inst/extdata/gencode.v47.annotation.gtf.gz \
  --genomeDir DOTSeq/inst/extdata/hg38_star_index

# Trim and align reads
for i in DOTSeq/inst/extdata/ly_2024/*/*/*.fastq.gz; do
  cutadapt -j 16 -m 15 -u 8 -e 0.1 --match-read-wildcards \
    -a TCGTATGCCGTCTTCTGCTTG -O 1 \
    -o $(dirname "$i")/$(basename "$i" .fastq.gz).trimmed.fasta.gz "$i"

  STAR --runMode alignReads \
    --runThreadN 32 \
    --outFilterType BySJout --outFilterMismatchNmax 2 \
    --genomeDir DOTSeq/inst/extdata/hg38_star_index \
    --readFilesIn $(dirname "$i")/$(basename "$i" .fastq.gz).trimmed.fasta.gz \
    --readFilesCommand zcat \
    --outFileNamePrefix $(dirname "$i")/$(basename "$i" .fastq.gz) \
    --outSAMtype BAM SortedByCoordinate \
    --quantMode TranscriptomeSAM GeneCounts \
    --outFilterMultimapNmax 1 --outFilterMatchNmin 16 \
    --alignEndsType EndToEnd --outSAMattributes All
done

Step 2. Prepare ORF-level GTF and BED files

Step-by-step on how to prepare ORF-level annotation is available in the vignettes. GTF files from GENCODE, Ensembl, or Araport should be used as input.

Alternatively, DOTSeq accept flattened annotation files generated using the RIBOSS engine.

# Generate ORF-level GTF using DOTSeq's Python script
python DOTSeq/inst/scripts/orf_to_gtf.py \
  --gtf DOTSeq/inst/extdata/gencode.v47.annotation.gtf.gz \
  --transcripts DOTSeq/inst/extdata/gencode.v47.pc_transcripts.fa.gz \
  --output DOTSeq/inst/extdata/dotseq

Step 3: Count reads

featureCounts -f -O -s 0 -T 16 \
-F GTF -a DOTSeq/inst/extdata/dotseq.gtf \
-o DOTSeq/inst/extdata/featureCounts.dotseq.out DOTSeq/inst/extdata/ly_2024/*/*/*Aligned.sortedByCoord.out.bam

Step 4: Run DOTSeq

Follow the analysis workflow as demonstrated in the vignettes.

CONTRIBUTING

We welcome contributions from the community! Whether it's fixing bugs, improving documentation, or suggesting new features, your input is valuable.

By participating in this project, you agree to abide by the terms outlined in this Contributor Code of Conduct.

To get started:

• Fork the repository
• Create a new branch for your feature or fix
• Submit a pull request with a clear description of your changes

If you have questions or ideas, feel free to open an issue or start a discussion.

CONTACTS AND BUG REPORTS

• Chun Shen Lim: chunshen [dot] lim [at] otago [dot] ac [dot] nz
• Gabrielle Chieng: gabrielle [dot] chieng [at] postgrad [dot] otago [dot] ac [dot] nz

CITATION

Chun Shen Lim, Gabrielle S. W. Chieng. (2026). DOTSeq enables genome-wide detection of differential ORF usage. BioRxiv. DOI: https://doi.org/10.1101/2025.09.24.678201