Skip to content


Switch branches/tags


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

GitHub Downloads Java CI with Maven

HaploGrep is a tool for mtDNA haplogroup classification. We provide HaploGrep as a fast and free haplogroup classification web service or as a commandline tool. You can upload your mtDNA profiles aligned to rCRS or RSRS (beta) and receive mitochondrial haplogroups in return. FASTA, VCF and hsd input files are supported. As of today (August 18, 2021), HaploGrep and the updated HaploGrep2 have been cited over 920 times (Google Scholar - August 18, 2021).

Download and Install

curl -sL | bash

If you want to use our web service, please click here.

Phylogenetic Trees

The haplogroup classifications in Haplogrep are based on the revised tree by Dür et al, 2021, which is an update of the latest PhyloTree version 17 by van Oven, 2016 based on the work of van Oven & Kayser, 2009.

Cite us

If you use HaploGrep, please cite our latest Haplogrep2 paper or the initial Haplogrep paper.

Additionally please cite (1) Dür et al, 2021 if you use the latest Phylotree17_FU1 tree, (2) van Oven, 2016 for PhyloTree 17 or van Oven & Kayser, 2009 in case an older PhyloTree version has been used.

Available Tools

Currently two tools are available.

  • Classify allows to classify input profiles (hsd, fasta, VCF) into haplogroups.
  • Distance calculates the distance between two haplogroups.

HaploGrep Classify

Run HaploGrep Classification with test data

# Download test data

# Run Haplogrep Classification
./haplogrep classify --in HG00097.vcf.gz --format vcf --out haplogroups.txt

Input File Formats

VCF or Fasta

The recommended input format is a single-sample/multi-sample VCF (*.vcf.gz or *.vcf).


For alignment, bwa version 0.7.17 is used. For each input sequence, HaploGrep excludes positions from the tested range that are (1) not covered by the input fragment or (2) has marked with a N in the sequence.

hsd Format

You can also specify your profiles in the original HaploGrep hsd format, which is a simple tab-delimited file format consisting of 4 columns (ID, Range, Haplogroup and Polymorphisms).

Sample1 1-16569 H100 263G 315.1C 750G	1041G	1438G	4769G	8860G	9410G	12358G	13656C	15326G	16189C	16192T	16519C
Sample2 1-16569 ? 73G	263G	315.1C 750G	1438G	3010A	3107C	4769G	5111T	8860G	10257T	12358G	15326G	16145A	16222T	16519C

For readability, the polymorphisms are also tab-delimited (so columns >= 4). A hsd example can be found here.

Required Parameters

Parameter Description
--in Please provide the input file name
--format Please provide the input format of your data - valid options are: hsd, vcf, or fasta files
--out Please provide an output name

Additional Parameters

Parameter Description
--rsrs By default HaploGrep expects that your data is aligned against rCRS (which is included in the human references hg19 and hg38). If your data is aligned against RSRS, add the --rsrs parameter (Default: off). Please read this blog post carefully before adding this option.
--metric To change the classification metric to Hamming Distance (hamming) or Jaccard (jaccard) add the --metric parameter (Default: Kulczynski Measure).
--extend-report For additional information on mtSNPs (e.g. found or remaining polymorphisms) please add the --extend-report flag (Default: off).
--phylotree The used Phylotree version can be changed using the --phylotree parameter (Default: 17_FU1, allowed numbers from 10,11,12,..,17 (latest version)).
--chip If you are using genotyping arrays, please add the --chip parameter to limit the range to array SNPs only (Default: off, VCF only). To get the same behaviour for hsd files, please add only the variants to the range, which are included on the array or in the range you have sequenced (e.g. control region). Range can be sepearted by a semicolon ;, both ranges and single positions are allowed (e.g. 16024-16569;1-576;8860).
--skip-alignment-rules Add this option to skip our rules that fixes the mtDNA nomenclature for fasta import. Click here for further information. Applying the rules is the default option since v2.4.0
--hits To export the best n hits for each sample add the --hits parameter. By default only the tophit is exported.
--lineage Create a graph of all input samples by using the --lineage parameter. (Default: 0). 0=no tree, 1=tree with genotypes, 2=only structure, no genotypes. As an output we provide a Graphviz DOT file. You can then use graphviz (sudo apt-get install graphviz) to convert the dot file to a e.g. pdf (dot <dot-file> -Tpdf > graph.pdf).

HaploGrep Distance

This tool allows to calculate the distance between two haplogroups.

Required Parameters

Parameter Description
--in Input file must include 2 columns named "hg1" and "hg2" seperated by ";"
--out Output location of distance file

mtDNA reference sequences

Several mtDNA references exist, HaploGrep supports rCRS and RSRS. Please checkout our blog post to learn more about this topic.

Genotyping arrays

If you are using HaploGrep for genotyping array data, please have a look at the --chip parameter above.

mtDNA Nomenclature

When using fasta as an input format, HaploGrep uses bwa mem to align data. Since the mitochondrial phylogeny is using a 3′ alignment, indels are often not correctly placed for haplogroup classification, when using standard-aligner designed for nuclear DNA. In some cases, where haplogroup defining indels are expected (e.g. missing 8281d-8289d) this can yield to a lower haplogroup quality. To adjust for that, we provide a set of currently 66 rules that can be applied prior to classification. The rules have been estimated based on 7,848 fasta files in 4 steps:

  1. Downloading Phylotree defining sequences from GenBank
  2. Aligning data with bwa mem
  3. Classifying the profiles using HaploGrep
  4. Comparing final fasta profiles with the Phylotree input profiles (remaining vs. not found) in a txt format (derived from parsing Phylotree). For example, the subsequent rule changes input polymorphisms 309.1CCT 310C to 309.1C 309.2C 315.1C.

Heteroplasmies (VCF only)

Heteroplasmies are often stored as heterozygous genotypes (0/1). If a AF tag (= Allele Frequency) is specified in the VCF file, we add variants with a AF > 0.90 to the input profile. Mutation Server is able to create a valid VCF including heteroplasmies starting from BAM or CRAM.

Related work

Please have a look at mitoverse to check for heteroplasmies and contamination in your NGS data.


Check out our blog regarding mtDNA topics.


Sebastian Schoenherr (@seppinho)
Hansi Weissensteiner (@haansi)
Institute of Genetic Epidemiology, Medical University of Innsbruck