This document summarizes a presentation on functional meta-analysis. It outlines the data sources used, which include gene expression arrays from studies of psoriasis and dermatitis and miRNA sequencing data from TCGA. It also describes the different meta-analysis models tested, including fixed effect and random effect models. The results sections shows the results of applying these models to identify significantly enriched pathways for psoriasis/dermatitis and differences between tumor and normal samples. It concludes that functional meta-analysis is an efficient method to confirm and discover functional profiles across different studies and functional annotations.
8. Data
Psoriasis and Dermatitis
○ Gene expression arrays
○ Human
○ Case vs. control
2
○ Without allergens
○ Selected platforms
P: 65
D: 36
P: 23
D: 18
16. Results3
method over under sig.over sig.under sig.ov.LOR sig.un.LOR
FE 842 878 312 387 63 87
HE 843 879 272 3151 51 80
HS 840 882 272 321 51 75
SJ 841 881 198 274 41 67
PM 841 881 263 311 51 75
DL 839 883 261 308 50 72
Level 1: tumor vs. normal Reactome pathways
1722 569 122
FDR < 0.05 | LOR | > 0.5
17. Results3
method over under sig.over sig.under sig.ov.LOR sig.un.LOR
FE 350 394 133 169 0 2
HE 350 394 120 136 0 2
HS 350 394 123 139 0 2
SJ 350 394 104 117 0 2
PM 350 394 120 136 0 2
DL 350 394 120 136 0 2
Level 1: tumor vs. normal Molecular Function
744 256 2
FDR < 0.05 | LOR | > 0.5
18. Conclusions4
Efficient method to confirm and discover
functional profiles
Flexible for functional annotations, high
throughput technologies data and methods
Assessment of Genomic Studies