Table S2: Confusion Matrix and out of bag (OOB) error rate of Random Forest classification of the fecal samples. Numbers on the diagonal of the table correspond to correctly classified samples while the off diagonal numbers represent misclassified samples. Rate of misclassification per class (treatment) is captured in the “class.error” column. 

Call:
 randomForest(x = x, y = y, ntree = 2000, mtry = 53, importance = TRUE) 
               Type of random forest: classification
                     Number of trees: 2000
No. of variables tried at each split: 53

        OOB estimate of  error rate: 20%
Confusion matrix:
     FliC IL1b MPER NG RB WT class.error
FliC   30    2    2  0  1  0   0.1428571
IL1b    0   27    6  1  0  1   0.2285714
MPER    0    6   25  0  0  1   0.2187500
NG      1    4    2 28  0  0   0.2000000
RB      4    2    1  0 32  0   0.1794872
WT      1    5    0  0  1 22   0.2413793