Author: Eyal Shahar
Position: Professor
Institution: The Univesrity of Arizona
E-mail: shahar@email.arizona.edu
Submitted Date: September 10, 2008
Published Date: September 12, 2008
This comment was originally posted as a “Reader Response” on the publication date indicated above. All Reader Responses are now available as comments.

I believe that in both Figure 3 and Figure 4, the marginal association between Gene 1 and disease is not a valid estimate of the effect of Gene 1 on disease. Gene 2 is a confounder in both cases. There is a back-door, unblocked path from Gene 1 to disease (Gene 1--Gene 2--Exposure--Disease), which creates a non-causal association between Gene 1 and Disease.
Although the segment "Exposure--> Disease" is part of that path, the full path is not a causal path for the effect of Gene 1 on Disease.

I would be interested in the authors' response.