Confusion matrices corresponding to the best gene and protein sets of the full datasets and the liver tissue validation datasets.

The Liver 3-way Full best gene and protein sets contained 33 genes and 27 proteins, respectively. The PBMC 3-Way Full best gene and protein sets contained 16 genes and 28 proteins, respectively. (A) Confusion matrix for classification of Liver 3-Way Full RNAseq dataset using best gene set identified by filter feature selection. The diagonal contains the number and percentage of the correctly predicted samples. (B) Confusion matrix for classification of AH, AC, and healthy control (CT) samples within independent validation RNAseq dataset. (C) Confusion matrix for classification of PBMC 3-Way Full RNAseq dataset using best gene set identified by filter feature selection. (D) Confusion matrix for classification of Liver 3-Way Full proteomic dataset using best protein set identified by filter feature selection. (E) Confusion matrix for classification of AH, AC, and CT samples within independent validation proteomic dataset. (F) Confusion matrix for classification of PBMC 3-Way Full proteomic dataset using best protein set identified by filter feature selection.