Fig. 1

The workflow of personalized health status monitoring framework. a Statistical modeling: Healthy samples from the discovery set are subjected to data preprocessing and subsequently used as a training set to train the model. The model establishes quantitatively a computational pipeline to infer the health index (\(\phi\)) and the BHC. b Health prediction: The health index and related threshold output from the PCA model are used to predict the health of the samples. c Further exploration: Health patterns discovery in the healthy population through contrastive PCA learning. In disease populations, the target diagnostic species and the contribution of species, determined by contribution analysis, identify broad-spectrum diagnostic species