Identifying temporal molecular signatures underlying cardiovascular diseases: A data science platform

Area of Science:

• Cardiovascular Science
• Bioinformatics
• Systems Biology

Background:

• Cardiovascular disease progression involves complex molecular changes in the heart.
• High-throughput omics technologies allow temporal profiling of biological systems.
• Computational analysis of temporal omics data, especially using unsupervised methods, presents significant challenges in cardiovascular research.

Purpose of the Study:

• To develop a computational platform for analyzing temporal omics datasets in cardiovascular research.
• To address the lack of specialized bioinformatic pipelines for unsupervised analysis of temporal cardiovascular data.
• To facilitate the extraction of biomedical insights from complex, time-resolved molecular data.

Main Methods:

• Developed a non-parametric data analysis platform (CV.Signature.TCP) with three modules.
• Module I: Preprocessing (cubic splines/PCA), missing data imputation, and denoising.
• Module II: Unsupervised clustering (K-means/hierarchical). Module III: Feature selection using jackstraw method for p-values and posterior inclusion probabilities (PIPs).

Main Results:

• Successfully applied the platform to a temporal proteomics dataset of oxidative stress-induced post-translational modifications (O-PTMs).
• Identified distinct temporal clusters and biological entities associated with specific temporal patterns in cardiac remodeling.
• Demonstrated the platform's utility in unveiling biological insights into O-PTM regulations during isoproterenol (ISO)-induced cardiac hypertrophy.

Conclusions:

• CV.Signature.TCP is an open-source R package for identifying temporal clusters in omics data.
• The platform effectively reveals biological insights in cardiovascular research, exemplified by cardiac remodeling studies.
• This tool enhances the analysis of temporal omics data, supporting the understanding of cardiovascular disease pathogenesis.