Gut microbiome disease signatures have been reported for a myriad of diseases including both non-communicable and communicable diseases. Collectively these findings beg the question of whether we can tell if a microbiome is healthy or not? However, commonalities among gut microbial signatures for different conditions have not been rigorously investigated. At the same time, our understanding of what constitutes a healthy microbiome is still imprecise – yet, many researchers freely use the terms eubiosis and dysbiosis to describe so-called ‘healthy’ and ‘diseased’ gut microbiome states. To gain a more quantitative understanding of dysbiosis and eubiosis, we will make use of public gut microbiome sequencing data and employ machine learning methodology to derive quantitative models of eubiosis and dysbiosis.