Fig. 1

The overall design of this study. A Schematic diagram for disease progression of a complex disease in a subject. Regardless of specific differences in either biological processes or observed symptoms among diseases, the progression of illness can be generally divided into three stages or states, i.e., a relatively normal (before-deterioration) state, a pre-disease (critical) state, and a disease (deteriorated) state. A relatively normal state and pre-disease state are reversible state but a disease state is irreversible state. Thus, detection of the critical state is essential; B one-sample based local network entropy algorithm. Given a number of reference samples which can be derived from normal cohort, the LNE is calculated based on a single-sample from any individual. Specifically, both the reference samples and the to-be-determined single-sample are mapped to the existing PPI network or other reference network, which can be partitioned into local networks. For each local network centered on gene k, the local LNE score \(\Delta E\left(t\right)\) is calculated; C for each cancer from the ten cancers and each individual from the cancer, calculating the LNE score of its each gene. After ranking the scores for all genes, the top 5% genes can be regarded as the LNE genes for the sample. LNE genes were further categorized to O-LNE and P-LNE biomarkers which enable significantly distinguish survival time between identified samples and unidentified samples