3  Build Salmon & STAR Index

3.1 Overview

Salmon and STAR are two cornerstone tools for RNA-seq quantification that serve complementary purposes:

  • Salmon — a fast, lightweight alignment-free (quasi-mapping / selective-alignment) quantifier that works on a transcriptome index, provides bias-aware transcript abundance estimates, and is ideal for rapid, large-scale quantification.
    Docs : Salmon Documentation · Github : COMBINE-lab/salmon

  • STAR — an alignment-based spliced read aligner that maps reads to the genome, discovers splice junctions, and produces alignment files suitable for downstream QC and gene-level counting.
    Github : alexdobin/STAR

This page shows how to build indices for both tools. Choose Salmon when you want fast transcript-level quantification from a transcriptome; choose STAR when you need full genomic alignments, splice junction discovery, or STAR-based counting workflows.

3.2 Prerequisites

  • Activate iobrpy environment
conda activate iobrpy
  • Choose a base directory for index:
export BASE=/path/to/index/dir
mkdir -p "$BASE"

3.3 Build Salmon index

# Move to the base directory for references
cd "$BASE"

# Download GENCODE v44 annotation & transcript FASTA
# (Use -c to resume if the download is interrupted)
wget -c ftp://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_44/gencode.v44.annotation.gtf.gz
wget -c ftp://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_44/gencode.v44.transcripts.fa.gz

# Decompress the downloads
gunzip -f gencode.v44.annotation.gtf.gz
gunzip -f gencode.v44.transcripts.fa.gz

# Paths
fa="$BASE/gencode.v44.transcripts.fa"

# Output directory for Salmon index
mkdir -p "$BASE/salmon"

# Build Salmon index
# -t: transcript FASTA
# -i: index output directory
# -k: k-mer size (31 is common for human)
# -p: threads
salmon index -t "$fa" -i "$BASE/salmon/gencode44" -k 31 -p 8

Tip: Adjust -p to match your CPU cores; -k can be tuned depending on read length and species.

3.4 Build STAR index

# Move to the base directory for references
cd "$BASE"

# Download the primary assembly and GTF (GENCODE v44)
wget -c https://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_44/GRCh38.primary_assembly.genome.fa.gz
wget -c https://ftp.ebi.ac.uk/pub/databases/gencode/Gencode_human/release_44/gencode.v44.annotation.gtf.gz

# Decompress the downloads
gunzip -f GRCh38.primary_assembly.genome.fa.gz
gunzip -f gencode.v44.annotation.gtf.gz

# Paths
STAR_INDEX="$BASE/star"
genome_fa="$BASE/GRCh38.primary_assembly.genome.fa"
gtf="$BASE/gencode.v44.annotation.gtf"

# Create output directory for STAR index
mkdir -p "$STAR_INDEX"

# Build STAR genome index
STAR --runMode genomeGenerate \
     --genomeDir "$STAR_INDEX" \
     --genomeFastaFiles "$genome_fa" \
     --sjdbGTFfile "$gtf" \
     --runThreadN 16 \
     --sjdbOverhang 100

3.5 Notes

  • Release/assembly: This guide uses GENCODE v44 (GRCh38). If you need a different release or assembly (e.g., T2T-CHM13), update the URLs and filenames accordingly.
  • Threads: Increase --runThreadN/-p to speed up indexing if you have more cores.
  • Storage: STAR indexes are large (tens of GB). Ensure you have enough disk space.
  • Read length: For STAR, set --sjdbOverhang = read_length - 1 for best splice junction annotations.

3.6 References & Acknowledgments

We gratefully acknowledge the developers and maintainers of Salmon and STAR. If you use these tools, please cite:

  • Patro R, Duggal G, Love MI, Irizarry RA, Kingsford C. Salmon provides fast and bias-aware quantification of transcript expression. Nat Methods. 2017 Apr;14(4):417-419. doi: 10.1038/nmeth.4197. Epub 2017 Mar 6. PMID: 28263959; PMCID: PMC5600148.

  • Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, Batut P, Chaisson M, Gingeras TR. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013 Jan 1;29(1):15-21. doi: 10.1093/bioinformatics/bts635. Epub 2012 Oct 25. PMID: 23104886; PMCID: PMC3530905.