Linkage Analysis of Gene Expression Traits
Author Information
Author(s): Li Na, Wu Baolin, Wei Peng, Xie Benhuai, Xie Yang, Xiao Guanghua, Pan Wei
Primary Institution: University of Minnesota
Hypothesis
Can combining multiple gene expression traits improve linkage analysis for complex diseases?
Conclusion
Using principal-components analysis (PCA) for gene expression traits generally provided stronger evidence for linkage than individual traits.
Supporting Evidence
- PCA generally yielded stronger evidence for linkage compared to individual traits.
- The LDA-derived trait had the worst linkage results despite having the highest heritability.
- The study identified ten functional groups based on heritability for linkage analysis.
Takeaway
The study looked at how to group gene expression traits to find links to diseases, and found that using a special math method called PCA worked better than looking at each trait alone.
Methodology
Transcripts were grouped by biological functions, and linkage analysis was performed using sample averages, PCA, and linear discriminant analysis (LDA).
Potential Biases
Potential bias due to the independence assumption of peak locations in p-value calculations.
Limitations
The study's p-value calculations were based on simplified assumptions, and the definition of functional groups may not imply direct regulatory relationships.
Statistical Information
P-Value
0.002 for 2 peaks, 3 × 10-5 for 3 peaks in Group 2; 0.01 for 2 peaks, 7 × 10-4 for 3 peaks in Group 5.
Statistical Significance
p<0.05
Want to read the original?
Access the complete publication on the publisher's website