3 ms·
Correlation, not causation. With 500-1000 identified/unidentified species of gut bacteria floating around, mimicking one is a bad data set. Corollary example,
by kumarski 12y ago
Correlation, not causation.
With 500-1000 identified/unidentified species of gut bacteria floating around, mimicking one is a bad data set.
Corollary example, we've not proved that H.Pylori causes Ulcerative Colitis, yet they coexist frequently.