Fig 5. Microbial species sorted by saliency for the Colorectal-EMBL dataset.
Saliency maps computed for the Colorectal-EMBL dataset using Eq 10. (A) Top 25 microbial species sorted by mean saliency. Species abundance significance in healthy (red) and affected (blue) individuals was calculated using a Wilcoxon test for each microbial species for two unpaired samples: healthy and affected individuals. Error bars represent the standard error over the five repeated experiments and the five ensembled MVIB models for each experiment. (B) Full distribution of the mean saliency across the microbial species of the Colorectal-EMBL dataset.