henhao
Sign in
BioinformaticsBioinformatics
Peer-reviewed journal article

STAR: ultrafast universal RNA-seq aligner

Alexander DobiniD, Carrie Davis, Felix J. Schlesinger, Jörg Drenkow, Chris Zaleski, Sonali Jha, Philippe J. BatutiD, Mark ChaissoniD, T GingerasiD

Abstract

MOTIVATION: Accurate alignment of high-throughput RNA-seq data is a challenging and yet unsolved problem because of the non-contiguous transcript structure, relatively short read lengths and constantly increasing throughput of the sequencing technologies. Currently available RNA-seq aligners suffer from high mapping error rates, low mapping speed, read length limitation and mapping biases. RESULTS: To align our large (>80 billon reads) ENCODE Transcriptome RNA-seq dataset, we developed the Spliced Transcripts Alignment to a Reference (STAR) software based on a previously undescribed RNA-seq alignment algorithm that uses sequential maximum mappable seed search in uncompressed suffix arrays followed by seed clustering and stitching procedure. STAR outperforms other aligners by a factor of >50 in mapping speed, aligning to the human genome 550 million 2 × 76 bp paired-end reads per hour on a modest 12-core server, while at the same time improving alignment sensitivity and precision. In addition to unbiased de novo detection of canonical junctions, STAR can discover non-canonical splices and chimeric (fusion) transcripts, and is also capable of mapping full-length RNA sequences. Using Roche 454 sequencing of reverse transcription polymerase chain reaction amplicons, we experimentally validated 1960 novel intergenic splice junctions with an 80-90% success rate, corroborating the high precision of the STAR mapping strategy. AVAILABILITY AND IMPLEMENTATION: STAR is implemented as a standalone C++ code. STAR is free open source software distributed under GPLv3 license and can be downloaded from http://code.google.com/p/rna-star/.

DOI
Structured summary ↓

02 /Specification & Classification

Click any tag to browse matching papers
Citations
51,628
FWCI
82.20
Field percentile
top 1%
OpenAlex Topics

Abstract summary

Verified abstract synthesis

Claude Opus 5.5 · max effort · 17 Oct 2026 · from the abstract

Maximum mappable seed search in uncompressed suffix arrays, then seed clustering and stitching, aligns 550 million paired-end reads per hour on 12 cores with better sensitivity and precision.

RNA-seq reads are short, spliced across exons and arriving in ever larger volumes, and existing aligners suffered from high mapping error rates, low speed, read-length limits and mapping biases. A seed-search-and-stitch algorithm built to process an ENCODE-scale dataset reports both a large speed gain and better alignment quality, while also finding non-canonical and chimeric junctions.

Key results

  1. ClaimedMapping speed exceeds other aligners by more than a factor of fifty while sensitivity and precision improve rather than degrade.

    550 million paired-end reads of 2 x 76 bp per hour against the human genome on a 12-core server; the competing aligners and the sensitivity and precision measurements are not named or quantified in the abstract. · Abstract

  2. ShownValidation by an orthogonal assay supports the precision claim for newly found junctions.

    1960 novel intergenic splice junctions tested by Roche 454 sequencing of reverse transcription PCR amplicons validated at an 80-90% success rate. · Abstract

  3. ClaimedCoverage goes beyond canonical junctions: de novo canonical junction detection plus non-canonical splices, chimeric or fusion transcripts, and full-length RNA sequences.

    Listed as capabilities in the abstract, with no per-category detection rates or benchmarks reported there. · Abstract

Methods, data and reproducibility

An RNA-seq alignment algorithm that performs sequential maximum mappable seed search against uncompressed suffix arrays, then clusters and stitches seeds into spliced alignments, implemented as standalone C++ software and developed to process an ENCODE transcriptome dataset.

  • ENCODE Transcriptome RNA-seq dataset that motivated the aligner · n = more than 80 billion reads
  • Paired-end reads of 2 x 76 bp aligned to the human genome for the speed measurement · human · n = 550 million reads per hour
  • Novel intergenic splice junctions tested by reverse transcription PCR amplicon sequencing · n = 1960 junctions
Methods & tools
STAR, Roche 454 sequencing, reverse transcription PCR, C++
Code
http://code.google.com/p/rna-star/
Software & environment
Standalone C++ code released as free open source under the GPLv3 license.
Assay conditions & compute
Speed benchmark run on a 12-core server; no memory footprint, processor model or suffix array build time is reported.
Limitations (2)
  • The speed advantage is given as a factor against unnamed aligners, and the claimed sensitivity and precision gains come without metrics or a benchmark description. · noted in this summary
  • Experimental validation covers intergenic junctions only, so precision for canonical junctions, chimeric transcripts and full-length RNA mapping is not independently confirmed. · noted in this summary

Who should read:Read this if you map large volumes of RNA-seq reads and need spliced alignment that also reports non-canonical junctions and fusion transcripts.

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISSarkar '17Srivastava '16Sahlin '21Kim '14Boratyn '19Shinder '23
72% semantic matchBioinformatics
Hirak Sarkar, Mohsen Zakeri, Laraib Malik+1 more·bioRxiv·Aug 24, 2017· 4 citations