Skip to main content
Fig. 3 | Microbiome

Fig. 3

From: Machine learning-assisted identification of bioindicators predicts medium-chain carboxylate production performance of an anaerobic mixed culture

Fig. 3

Random forest feature importance of ASVs used to classify the HRT phases (A-HRT bioindicators and B-HRT bioindicators). The top-ranked 15 ASVs reducing the uncertainty in the prediction of HRT phases (HRT of 8 days and 2 days). According to their ASV abundances distribution, the order of features (from top to bottom) was based on their mean decrease in Gini scores, with HRT as the response variable. a Feature importance of A-HRT bioindicators. The ASV importance was calculated using the relative abundance data of bioreactor A as a training set and data of bioreactor B as a test set. b Feature importance of B-HRT bioindicators. Similar to A-HRT bioindicators, ASV importance of B-HRT was calculated using the relative abundance data of bioreactor B as a training set and data of bioreactor A as a test set. The taxonomic classification of ASVs assigned at the genus level is provided in parentheses

Back to article page

Microbiome

ISSN: 2049-2618

Contact us