MDR


Welcome to the new homepage of the Multifactor Dimensionality Reduction (MDR) open-source software package.

MDR is a data mining strategy for detecting and characterizing nonlinear interactions among discrete attributes (e.g. SNPs, smoking, gender, etc.) that are predictive of a discrete outcome (e.g. case-control status).  The MDR software combines attribute selection, attribute construction and classification with cross-validation to provide a powerful approach to modeling interactions.

Click here to download the latest version of the free open-source MDR software.

See Epistasis Blog for an MDR 101 tutorial (Nov. 12 - Dec. 4, 2006).

See the Wikipedia entry for a conceptual introduction to MDR.

See our recent paper in the Journal of Theoretical Biology for a scientific review of MDR.

Click here to see a list of papers I know about that have applied MDR to real data.

Click here to carry out a PubMed search for MDR papers.

Click here to Google MDR.

Click here to Google Scholar MDR.

For more information about MDR please see Epistasis.org or Epistasis Blog.

Development and evaluation of MDR is funded by NIH grants AI59694, HD047447, and LM009012 and by generous funds from the Norris-Cotton Cancer Center at Dartmouth Medical School.

Send questions or feedback to:



Here are some screenshots:





Page last updated by JHM January 7, 2007