FILTUS
FILTUS is a stand-alone tool for downstream
analysis in high-throughput sequencing projects. It is especially well
suited for identification of variants causing Mendelian disease. The program is written in Python, by:
Magnus Dehli Vigeland, PhD
Department of Medical Genetics
Oslo University Hospital, Norway
Email: magnusdv<at>medisin<dot>uio<dot>no
FILTUS accepts (at least in principle) any format of variant files, as
long as you state which columns contain the essential
information (e.g. chromosome, position, gene and genotype). Standard
VCF files are automatically recognized. Unlike other variant tools, FILTUS is completely self-contained and
does not require an internet connection. This may be an important point
if you are not allowed to be online when working with human
sequence data.
The latest version of FILTUS is 1.0.5. Take a
look at the changelog
to see recent developments. Questions,
bug reports, feature requests and other comments are welcome by email
to magnusdv<at>medisin<dot>uio<dot>no.
Features
- Overview statistics of files and of individual columns.
- Filters:
- Any number of column filters
- Exclude variants in inhouse datase
- Exclude/restrict to genes in a gene list
- Restrict to certain genomic regions
- Breakdown table of the filtering process.
- De novo analysis in trios.
- Autozygosity mapping in individuals with known or suspected inbreeding ("AutEx").
- Gene based analysis: Identify genes where specified samples
contain variants.
- Genetic models: Dominant, simple recessive, compound
heterozygous.
- Statistical model for prioritizing genes in Mendelian
disease mapping projects.
- Sporadic patients or family based analysis.
- Export to MERLIN-format (for linkage analysis).
- Simultaneous analysis of up to several hundred complete
exomes.
- Prefilter files when loading: Allows analysis of whole-genome size files.
- Quality control plots:
- Compare heterozygosity levels, private variants and other
features.
- Scatter plots of any numerical columns
- stratified on genotype.
- Reproducibility: Save/load filters.
Download and installation
WINDOWS
- Download the latest version here
and unpack the zip-file in any location you prefer. To start the
program, simply double-click on the file FiltusXXX.exe in the
unpacked folder (XXX = version number).
A couple of things to notice:
- The windows version of FILTUS requires a 64-bit processor.
- The unpacked Filtus folder contains a great number of
files in addition to the .exe file. These are for internal use only and
should be ignored (but not deleted) by end users.
- To create a FILTUS icon on the desktop (Windows 7):
- Right click on the .exe file
- Choose Send to and then Desktop
(create shortcut)
- To uninstall FILTUS you can just delete the Filtus folder.
FILTUS leaves no footprint on your computer.
MAC/LINUX
- If you have pip installed, automatic installation can be done by pip install filtus in a terminal.
- Alternatively, you are welcome to download the source files
and run Filtus from a command line:
- Download the source files here and unpack
the zip-file in a location of your choice.
- Open a terminal and move to the unpacked Filtus folder.
- To start the program, use the command python Filtus.py
Requirements when running FILTUS from source files:
- Python version 2.7.x where x is at least 3. (Get
it from python.org
if you don't have it.) In particular, Python 3 will not work.
- For the plotting functionality (QC plots) to work, the two
Python packages numpy
and matplotlib
must be installed on your system.
Loading a test file
To check that everything is OK, you can load a test file as follows:
- Start FILTUS.
- On the File menu,
choose Load files (simple).
- Move to the testfiles
folder, choose test_file.csv and
press Open.
- In the Input file settings dialog, click Use
for all files.
- You should now see the file name in the Loaded
samples window, and a row in the Unfiltered
summaries window saying "425 variants in 205 genes". If you
double-click on this row, you should see the contents of the file in
the main window.