Detection for disease tipping points by landscape dynamic network biomarkers

Area of Science:

• Biomedical Engineering
• Computational Biology
• Systems Biology

Background:

• Early disease detection is crucial for effective intervention.
• Current methods often require longitudinal data or are not sensitive to individual patient states.
• Identifying biomarkers that predict disease transition is a significant challenge.

Purpose of the Study:

• To introduce and validate a novel model-free methodology, landscape dynamic network biomarker (l-DNB), for early disease prediction.
• To demonstrate the capability of l-DNB in identifying critical genes and network biomarkers associated with disease progression.
• To apply l-DNB to predict severe influenza and analyze tumor deterioration using TCGA datasets.

Main Methods:

• Developed the landscape dynamic network biomarker (l-DNB) methodology based on bifurcation theory.
• Applied l-DNB to single-sample omics data for early warning signal detection.
• Utilized l-DNB to identify critical genes (DNB members) driving disease transitions.
• Validated l-DNB on influenza virus infection data and three TCGA tumor datasets.

Main Results:

• l-DNB successfully provided early-warning signals for disease deterioration on a single-sample basis.
• Identified critical genes and network biomarkers (DNB members) that promote the transition from normal to disease states.
• Predicted severe influenza symptoms prior to clinical manifestation.
• Detected critical stages of tumor deterioration and identified two types of prognostic biomarkers for cancer.

Conclusions:

• The landscape dynamic network biomarker (l-DNB) methodology offers a powerful tool for early disease detection and biomarker discovery using single-sample omics data.
• l-DNB can predict disease deterioration and identify key molecular drivers, applicable across various diseases including infections and cancers.
• The identified prognostic biomarkers show potential as common biomarkers for predicting cancer patient outcomes.