Skip to content
Switch branches/tags


Failed to load latest commit information.
Latest commit message
Commit time


A fast structural variant caller for long-read sequencing, Sniffles2 accurately detect SVs on germline, somatic and population-level for PacBio and Oxford Nanopore read data.

Quick Start: Germline SV calling using Sniffles2

To call SVs from long read alignments (PacBio / ONT), you can use:

sniffles -i mapped_input.bam -v output.vcf

For improved calling in repetitive regions, Sniffles2 accepts a tandem repeat annotations file using the option --tandem-repeats annotations.bed. Sniffles2 compatible tandem repeat annotations for human references can be downloaded from the annotations/ folder.

(see sniffles --help or below for full usage information).


You can install Sniffles2 using pip or conda using:

pip install sniffles


conda install sniffles=2.0

If you previously installed Sniffles1 using conda and want to upgrade to Sniffles2, you can use:

conda update sniffles=2.0


  • Python >= 3.7
  • pysam

Tested on:

  • python==3.9.5
  • pysam==


Please cite our paper at:

A new preprint for the new methods and improvements introduced with Sniffles2 is here:

Use-Cases / Modes

A. General (all Modes)

  • To output deletion (DEL SV) sequences, the reference genome (.fasta) must be specified using e.g. --reference reference.fasta.
  • Sniffles2 supports optionally specifying tandem repeat region annotations (.bed), which can improve calling in these regions --tandem-repeats annotations.bed. Sniffles2 compatible tandem repeat annotations for human references can be found in the annotations/ folder.
  • Sniffles2 is fully parallelized and uses 4 threads by default. This value can be adapted using e.g. --threads 4 as option. Memory requirements will increase with the number of threads used.
  • To output read names in SNF and VCF files, the --output-rnames option is required.

B. Multi-Sample SV Calling (Trios, Populations)

Multi-sample SV calling using Sniffles2 population mode works in two steps:

  1. Call SV candidates and create an associated .snf file for each sample: sniffles --input sample1.bam --snf sample1.snf
  2. Combined calling using multiple .snf files into a single .vcf: sniffles --input sample1.snf sample2.snf ... sampleN.snf --vcf multisample.vcf

Alternatively, for step 2. you can supply a .tsv file, containing a list of .snf files, and custom sample ids in an optional second column (one sample per line), .e.g.: 2. Combined calling using a .tsv as sample list: sniffles --input snf_files_list.tsv --vcf multisample.vcf

C. Non-Germline SV Calling (Somatic)

To call non-germline SVs (i.e. somatic/mosaic) SVs, the --non-germline option should be added, i.e.:

sniffles --input mapped_input.bam --vcf output.vcf --non-germline

D. Genotyping a known set of SVs (Force Calling)

Example command, to determine the genotype of each SV in input_known_svs.vcf for sample.bam and write the re-genotyped SVs to output_genotypes.vcf:

sniffles --input sample.bam --genotype-vcf input_known_svs.vcf --vcf output_genotypes.vcf

Quick Tips

Input / Output

  • .bam or .cram files containing long read alignments (i.e. from minimap2 or ngmlr) are supported as input
  • .vcf.gz (bgzipped+tabix indexed) output is supported
  • Simultaneous output of both .vcf and .snf file (for multi-sample calling) is supported