DAG program

released May 2011

Confounding is an important source of bias in epidemiologic studies. With the introduction of causal diagrams (directed acyclic graphs, DAG) a new approach to conceptualize confounding and new rules to identify the minimal sufficient adjustment set have been established (Greenland, Pearl & Robins, Epidemiology, 1999, 10(1):37-48).

The DAG program is a analysis tool, designed to select minimal sufficient adjustment sets within directed acyclic graphs.

The first version (v0.12) of the DAG program can be used as MS DOS command line program. Using the framework Qt (http://qt.nokia.com) we have developed a GUI that allows analyzing and changing causal diagrams more easier.

If you use DAG program (v0.21), you need the same inputfile as they used in the previos version. To learn more about the DAG program, you should read program documentation.

To visualize causal diagrams we refer to the webtool DAGitty (http://www.dagitty.net) developed by Johannes Textor.

Please note that we continuously improve the DAG program. If you have any problems, would like to report a bug or comment on the DAG program, please send an e-mail to sven.knueppel@dife.de.


DAG program is available for free download.

Platform Description File Size
WINDOWS Version v0.21 (May 12, 2011) dag_v021.exe 9194 KB
WINDOWS Qt source v0.21 (.zip) dag_v021.zip 37 KB
WINDOWS Program documentation DAG program v0.21 Tutorial.pdf 4738 KB
MS DOS Changelog Changelog.txt 1 KB
Old version
WINDOWS Qt source v0.20 (Nov 11, 2010) dag_v020.zip 36 KB
MS DOS Version v0.12 (March 31, 2010) dag.exe 1221 KB
MS DOS Program documentation DAG v0.12 PDF documentation.pdf 114 KB
MS DOS Example files example.zip 2 KB
MS DOS How to use DAG program v0.12? DAG program v0.12 usage.pdf 38 KB


Developed by

Sven Knüppel
Deutsches Institut für Ernährungsforschung Potsdam-Rehbrücke
German Institute of Human Nutrition Potsdam-Rehbruecke
Abteilung Epidemiologie
Arthur-Scheunert-Allee 114-116
14558 Nuthetal, Germany

e-mail: sven.knueppel@dife.de

Many thanks to Andreas Stang (Martin-Luther-University of Halle-Wittenberg, Germany) introducing me to the world of causal diagrams and for his assistant on the first version of the DAG program.

I would like to thank a number of people for helpful discussions that clarified my thinking on causal diagrams and for their comments on DAG program: Johannes Textor, Juliane Hardt, Karina Meidtner, Dagmar Drogan and especially Charles Poole.

This program comes with ABSOLUTELY NO WARRANTY; This is free software, and you are welcome to redistribute it under certain conditions (GNU GENERAL PUBLIC LICENSE (GPL), Version 3 or later).