Table 4.

Performance of random forest models built using glycomics, metallomics, metabolomics, lipidomics, and multi-omics features in classifying between AMI and healthy.

Classifier	No. of Features Used	AUCROC	CI
Glycomics	37	0.786	0.688–0.883
Metallomics	30	0.851	0.782–0.904
Metabolomics	27	0.836	0.744–0.930
Lipidomics	48	0.822	0.724–0.905
Multi-omics	Top 100 out of 142	0.953	0.911–0.987