Inflammatory bowel disease (IBD) is an autoimmune condition that is difficult to diagnose, and animal models of this disease have questionable human relevance1. Here, we show that the dysbiosis network underlying IBD in dogs differs from that in humans, with some bacteria such as Fusobacterium switching roles between the two species (as Bacteroides fragilis switches roles between humans and mice)2. For example, a dysbiosis index trained on humans fails when applied to dogs, but a dog-specific dysbiosis index achieves high correlations with the overall dog microbial community diversity patterns. In addition, a random forest classifier trained on dog-specific samples achieves high discriminatory power, even when using stool samples rather than the mucosal biopsies required for high discriminatory power in humans2. These relationships were not detected in previously published dog IBD data sets due to their limited sample size and statistical power3. Taken together, these results reveal the need to train host-specific dysbiosis networks and point the way towards a generalized understanding of IBD across different mammalian models.