Quick start

This walks through the minimum needed to run rustar-aligner end-to-end. If you already use STAR, the commands will look familiar — they are.

1. Generate a genome index

You need to do this once per reference genome.

rustar-aligner --runMode genomeGenerate \
  --genomeDir /path/to/genome_index \
  --genomeFastaFiles /path/to/genome.fa

For a human genome you’ll typically also pass a GTF file and tune the SA index parameter:

rustar-aligner --runMode genomeGenerate \
  --runThreadN 16 \
  --genomeDir /path/to/genome_index \
  --genomeFastaFiles /path/to/GRCh38.fa \
  --sjdbGTFfile /path/to/gencode.gtf \
  --sjdbOverhang 100

See the genome index guide for the full set of relevant parameters.

2. Align reads (single-end)

rustar-aligner \
  --genomeDir /path/to/genome_index \
  --readFilesIn reads.fq \
  --outSAMtype SAM \
  --outSAMstrandField intronMotif \
  --outFileNamePrefix sample_

This writes sample_Aligned.out.sam plus sample_Log.final.out, sample_SJ.out.tab, and other STAR-style output files into the current directory. Override the destination with --outFileNamePrefix /path/to/sample_.

3. Align reads (paired-end)

Pass two files to --readFilesIn:

rustar-aligner \
  --genomeDir /path/to/genome_index \
  --readFilesIn reads_1.fq reads_2.fq \
  --outSAMtype SAM \
  --outFileNamePrefix sample_

4. Get a sorted BAM directly

rustar-aligner \
  --genomeDir /path/to/genome_index \
  --readFilesIn reads.fq \
  --outSAMtype BAM SortedByCoordinate \
  --outFileNamePrefix sample_

The output is sample_Aligned.sortedByCoord.out.bam. Pass BAM Unsorted instead for an unsorted BAM.

5. Gzipped FASTQ input

Use --readFilesCommand zcat to decompress on the fly:

rustar-aligner \
  --genomeDir /path/to/genome_index \
  --readFilesIn reads_1.fq.gz reads_2.fq.gz \
  --readFilesCommand zcat \
  --outSAMtype BAM SortedByCoordinate \
  --outFileNamePrefix sample_

6. Inspect the output

After a successful run you get the standard STAR file set:

File	Contents
*_Aligned.out.sam / *_Aligned.sortedByCoord.out.bam	The alignments
*_Log.final.out	Summary statistics (MultiQC-compatible)
*_Log.out	Verbose run log
*_Log.progress.out	Per-chunk progress
*_SJ.out.tab	Splice junctions discovered during alignment

See the output files reference for a full breakdown of each file.

Common next steps

• Two-pass alignment for better novel-junction recovery: see the two-pass guide.
• Gene-level counts for differential expression: see the quantification guide.
• Chimeric detection for fusion calling: see the chimeric guide.