Survival analysis of longitudinal microarrays

Area of Science:

• Genomics
• Biostatistics
• Bioinformatics

Background:

• Linking gene expression to clinical and phenotypic characteristics is crucial in genomic research.
• Existing methods effectively relate gene expression to categorized or continuous data, but less so to event times (e.g., survival, relapse).
• There is a need for methods to analyze survival with longitudinally collected gene expression data.

Purpose of the Study:

• To develop and present a novel approach for the survival analysis of longitudinal gene expression data.
• To establish a measure of association between event time and time-varying gene expressions.
• To apply and validate the proposed method on a real-world dataset.

Main Methods:

• Construction of a novel measure of association for survival analysis with longitudinal gene expression.
• Utilizing permutation tests for statistical significance assessment.
• Implementing false discovery rate control for robust findings.

Main Results:

• The study successfully illustrates an approach for survival analysis of longitudinal gene expression data.
• A measure of association was constructed to link time to an event with gene expressions collected over time.
• The method was applied to a multi-center study on inflammation and response to injury.

Conclusions:

• The proposed method provides a framework for analyzing survival data alongside longitudinal gene expression.
• This approach can help uncover biological drivers of differential patient outcomes.
• Further application of this method can enhance understanding of gene function in clinical contexts.