My beta analysis (Bray-Curtis, weighted and unweighted distance) did not show any separation between conditions (15 samples control and 15 treated) while Jaccard distance showed clear separation between groups. Kindly, what is best possible explanation for these results?
alpha and PERMANOVA analysis did not show significant differences as well.
There’s a really nice description of the different diversity metrics. Jaccard distance accounts for the presence or absence of each features in calculating the distance, giving equal weight to each OTU. Unweighted UniFrac considers the evolutionary similarity. So, you may have individual features that are different but are closely related.