Fig. 4: Metrics on the Adenocarcinoma dataset. | Nature Communications

Fig. 4: Metrics on the Adenocarcinoma dataset.

From: BERNN: Enhancing classification of Liquid Chromatography Mass Spectrometry data with batch effect removal neural networks

Fig. 4

A Valid and test MCC scores for all methods benchmarked. Higher is better. The conditions compared are colorectal cancer and chronic enteritis. B Batch mixing metrics: normalized Batch Entropy (nBE), Adjusted Rand Index (ARI), Adjusted Mutual Information (AMI). Smaller is better. C QC metrics: Normalized Median Euclidean distance (QC nMED) and QC average Pearson Correlation Coefficient (qc_aPCC). Lower nMED and 1-qc_aPCC is better. MCC is compared to D nBE, E ARI, F QC nMED and G QC aPCC. Error bars represent standard deviations around the means. All error bars are derived from the results of 3-fold cross-validation (n = 3). The BERNN models are underlined. Source data are provided as a Source Data file.

Back to article page