DNAvi is an open-source software to integrate, analyze and visualize multiple cell-free DNA fragmentation profiles from liquid biopsies. It uses either gel images or automated gel electrophoresis device output tables (electropherograms). The tool was tested on a 32 GB memory local PC on Ubuntu 24.04.2 LTS, as well as on Windows 10 and macOS.
pip install dnavi
Done! You can move to Quick start.
Please make sure you have installed python ≥ 3.12. Next, download the required packages:
Python packages:
pip install numpy pandas seaborn scipy matplotlib imageio scikit-image werkzeug scikit-posthocs
Conda:
conda create --name dnavi numpy pandas seaborn scipy matplotlib imageio werkzeug scikit-image -y
conda activate dnavi
conda install conda-forge::scikit-posthocs -y
Next, download the repository.
Through https:
git clone https://github.com/anjahess/DNAvi.git
Through github CLI:
gh repo clone anjahess/DNAvi
Or through zip download:
Go to <Code> (top right), then click 'Download ZIP'.
Unpack or move the DNAvi folder to your location of choice and you're ready to start.
Linux: Ctrl+Alt+T
macOS: Launchpad -> Search Terminal -> Click on Terminal
Windows: Windows Symbol -> search cmd.exe -> Click cmd.exe
In this example we will run DNAvi on a test electropherogram signal table provided in this package:
cd path/to/DNAvi/
Hint: to find where DNAvi is installed type "pip show dnavi".
dnavi -i tests/electropherogram.csv -l tests/ladder.csv -m tests/metadata.csv
Or, if you installed from github:
python3 DNAvi.py -i tests/electropherogram.csv -l tests/ladder.csv -m tests/metadata.csv
This will result in the following output:
Welcome to
____ _ _ _ _
| _ | \ | | / \__ _(_)
| | | | \| | / _ \ \ / / |
| |_| | |\ |/ ___ \ V /| |
|____/|_| \_/_/ \_\_/ |_|
DNA file: tests/electropherogram.csv
Ladder file: tests/ladder.csv
Meta file: tests/metadata.csv
------------------------------------------------------------
ELECTROPHEROGRAM DNA SIZE ANALYSIS
------------------------------------------------------------
DNA file: tests/electropherogram.csv
Ladder file: tests/ladder.csv
Meta file: None
Saving results to: tests/results/
------------------------------------------------------------
... and additional infos depending on the analysis details. Once DNAvi is finished, this message will appear:
You can now go to you results folder:
├── results
├── plots
├── stats
└── qc
In the plots and stats folder, you will find various visualizations summarizing fragmentomic traces of your samples.
If you have multiple gel images or csv files to process, just put them into a folder and point DNAvi to that folder:
! Attention: Run together only files that have the same DNA ladder.
dnavi -i /path/to/folder -l ladder.csv
If you need help, simply run
dnavi --help
Which will result in a display of command line arguments with additional explanaitons:
Welcome to
____ _ _ _ _
| _ | \ | | / \__ _(_)
| | | | \| | / _ \ \ / / |
| |_| | |\ |/ ___ \ V /| |
|____/|_| \_/_/ \_\_/ |_|
usage: DNAvi.py [-h] [-i [<input-file-or-folder>]] -l [<ladder-file>] [-m [<metadata-file>]] [-n [<run-name>]] [-c [<config-file>]] [-iv [<(start,step)>]]
[-p] [-un] [-nt [<sample_name>]] [-ml <int>] [-incl] [-cor] [--verbose] [-v]
Analyse Electropherogram data e.g. for cell-free DNA from liquid biopsies
options:
-h, --help show this help message and exit
-i [<input-file-or-folder>], --input [<input-file-or-folder>]
Path to electropherogram table file or image file OR directory containing those files. Accepted formats: .csv/.png/.jpeg/.jpg or
directory containing those.
-l [<ladder-file>], --ladder [<ladder-file>]
Path to ladder table file. Accepted format: .csv
-m [<metadata-file>], --meta [<metadata-file>]
Path to metadata table file containing grouping information for input file (e.g. age, sex, disease). Accepted format: .csv
-n [<run-name>], --name [<run-name>]
Name of your run/experiment. Will define output folder name
-c [<config-file>], --config [<config-file>]
Define nucleosomal fractions with this path to a configuration file containing custom (nucleosome) intervals for statistics.
Accepted format: tab-separated text files (.txt)
-iv [<(start,step)>], --interval [<(start,step)>]
Auto-generate nucleosomal size intervals by providing (start,step), e.g. start at 100 and increase by 200 bp
-p, --paired Perform paired statistical testing
-un, --unnormalized Do not perform min/max normalization. ATTENTION: will be DNA-concentration sensitive.
-nt [<sample_name>], --normalize_to [<sample_name>]
Name of the sample to normalize all values to. ATTENTION: will be DNA-concentration sensitive.
-ml <int>, --marker_lane <int>
Change the lane selected as the DNA marker/ladder, default is first lane (1). Using this will force to use the specified column
even if other columns are called Ladder already.
-incl, --include Include marker bands into analysis and plotting.
-cor, --correct Perform advanced automatic marker lane detection in samples with highly variant concentrations (e.g., dilution series), so that
the marker borders will be determined for each sample individually
--verbose increase output verbosity
-v, --version show program's version number and exit
Special thanks to Yara Matani and Philine Guckelberger for testing DNAvi!
Source data to the publication below is available in this repository in the sourcedata directory.
Please cite DNAvi as: Hess, A., Seelow, D. & Kretzmer, H. DNAvi: Integration, statistics, and visualization of cell-free DNA fragment traces. Bioinformatics (2026) doi:10.1093/bioinformatics/btag041.