Skip to content

Kallisto#

Kallisto is a lightweight, ultra-fast RNA-seq quantification tool. It is mainly used for estimating transcript abundances directly from RNA-seq reads without performing full read alignment. A detailed tutorial is available here.

Key features#

  • Alignment-free: Kallisto does not align reads base-by-base to the genome. Instead, it uses a technique called pseudoalignment to quickly determine which transcript reads are compatible with.
  • Extremely fast: It is significantly faster than traditional aligners like STAR and HISAT2 because it skips full alignment.
  • Quantification: Directly outputs transcript abundance estimates such as TPM (Transcripts Per Million).
  • Low memory usage: Efficient enough to run on laptops or small servers.

Download#

Kallisto is already installed on H4H. If you would like to install it on your local computer, please follow these instructions.

Workflow:#

  • Step 1: Build a transcriptome index (from a reference transcriptome FASTA file) 
    kallisto index -i transcriptome.idx transcripts.fa
  • Step 2: Quantify RNA-seq reads (Pseudoalignment and abundance estimation) 
    kallisto quant -i transcriptome.idx -o output_dir -b 100 reads_1.fastq reads_2.fastq
    The option -b 100 specifies 100 bootstrap samples for estimating quantification uncertainty.

Outputs#

  • abundance.tsv — Main quantification table (TPM, estimated counts)
  • abundance.h5 - Binary HDF5 file containing the same information as abundance.tsv, plus bootstrap results if bootstrapping was performed.
  • run_info.json — Summary of the Kallisto run (e.g., number of processed reads, parameters used).

Usage#

  • A Snakemake Kallisto pipeline is available at H4H: /cluster/projects/bhklab/pipelines/kallisto_snakemake_pipeline
  • Details about pipeline setup and usage are available in the pipeline's README file.