Reconstructing disease dynamics for mechanistic insights and clinical benefit

Area of Science:

• Computational biology and bioinformatics
• Cancer research and molecular oncology
• Systems biology and disease dynamics

Background:

• Understanding disease progression is crucial for effective diagnostics and treatment.
• Disease dynamics are complex and highly heterogeneous between individuals, making them difficult to capture.
• Current methods struggle to analyze high-dimensional, short time-series data for disease progression.

Purpose of the Study:

• To present TimeAx, a novel algorithm for capturing disease progression dynamics.
• To provide a comparative framework for analyzing high-dimensional, short time-series data.
• To demonstrate the clinical utility of TimeAx in disease stratification and outcome prediction.

Main Methods:

• Development of the TimeAx algorithm for comparative disease dynamics analysis.
• Application of TimeAx to high-dimensional, short time-series data from multiple diseases.
• Analysis of urothelial bladder cancer progression, including molecular and microenvironmental factors.

Main Results:

• Identification of a stromal pro-invasion point in urothelial bladder cancer, linked to immune infiltration and mortality.
• TimeAx model differentiates early and late tumors within subtypes, revealing molecular transitions.
• Uncovered potential targetable pathways and improved molecular interpretability of disease progression.

Conclusions:

• TimeAx offers a powerful approach for studying complex disease progression dynamics.
• The algorithm enhances molecular interpretability and provides clinical benefits for patient stratification.
• TimeAx facilitates improved outcome prediction and identification of therapeutic targets.