Longitudinal Analysis of Contrasts in Gene Expression Data.

Area of Science:

• Biostatistics
• Genomics
• Computational Biology

Background:

• Longitudinal analysis is crucial for understanding disease progression.
• Multiple organ dysfunction syndrome (MODS) requires sensitive methods for detecting biological changes.
• Gene expression profiling provides insights into cellular responses.

Purpose of the Study:

• To develop a statistical testing methodology for detecting baseline departures in longitudinal gene expression data.
• To identify genes with differential intercepts between two patient groups (A and B) in the context of MODS.
• To apply this methodology to gene expression data from individuals with MODS.

Main Methods:

• Calculating gene expression contrasts between two time points for each individual and gene.
• Performing linear regression of gene expression contrasts on individual age, analyzed per gene.
• Developing hypothesis testing for detecting differences in regression intercepts between groups A and B.

Main Results:

• A novel testing methodology was developed for longitudinal gene expression analysis.
• The method effectively identifies genes with distinct baseline expression trajectories between patient groups.
• The approach was validated using a bootstrapped dataset derived from a MODS study.

Conclusions:

• The developed methodology provides a robust way to detect group-specific gene expression changes over time.
• This approach can aid in understanding the molecular mechanisms underlying MODS.
• The findings support the use of hypothesis testing on linear regression intercepts for differential gene expression analysis in longitudinal studies.