Extended Data Fig. 4: FUGAsseM accurately predicts annotations validated by experimental evidence.
From: Predicting functions of uncharacterized gene products from microbial communities
(a,b) The FUGAsseM-full (a) and FUGAsseM-MTX (b) models retained high accuracy when predicting annotations for experimentally validated annotations from the gold standard dataset (n = 142 and 285 total terms evaluated with and without experimental evidence, respectively). Box plots display the median (line at the 50th percentile), interquartile range (box spanning the 25th to 75th percentiles), whiskers (extending to 1.5× IQR), and mean values (dark points). (c,d) Distribution of Random Forest importance scores of FUGAsseM-full model’s second (data integration) layer, which predicted annotations with (c) (n = 3,725 term-species pairs for prediction) and without (d) (n = 11,513 term-species pairs for prediction) experimental time in the gold standard set. Box plots as in a.
