Table 2.
Analysis of overlapping LXR- and PPARG-responsive genes
| Overlap 24 h (%) | Concordant 24 h | Discordant 24 h | Percentage concordant 24 h | Overlap 48 h (%) | Concordant 24 h | Discordant 24 h | Percentage concordant 48 h | |
|---|---|---|---|---|---|---|---|---|
| GW3965 (fc ±2) | 32 (25.8 %) | 29 | 3 | 90.6 % | 203 (18.2 %) | 119 | 84 | 58.6 % |
| T0901317 (fc ±2) | 9 (15.5 %) | 8 | 1 | 88.9 % | 189 (16.9 %) | 70 | 119 | 37.0 % |
| GW3965 (fc ±1.5) | 119 (38.9 %) | 100 | 19 | 84.0 % | 790 (35.1 %) | 381 | 409 | 48.2 % |
| T0901317 (fc ±1.5) | 60 (35.1 %) | 41 | 19 | 68.3 % | 775 (32.1 %) | 222 | 553 | 28.6 % |
The overlap between LXR-responsive genes (from GW3965 and T0901317 treatments) with PPARG-responsive genes (from rosiglitazone treatment) is shown at both 24 and 48 h. The number and percentage of total responsive genes is shown. The number of genes and percentages that show concordant and opposing changes in differential expression across all drug comparisons is also given. Hypergeometric tests of gene overlaps for both drug comparisons and time points were highly significant (p < 1e-16)