Table 3.
Summary of direction of effects for genes with both cis- and trans-eQTL
| Treatment: drought | Treatment: recovery | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Hotspot ID | n. | Coordinates (Mb) | n. reinforce | n. antag. | Odds v. background | Odds v. neutral | n. reinforce. | n. antag. | Odds v. background | Odds v. neutral |
| 3a | 206 | 0.1–7.3 | 43 | 42 | 2.27** | 1.02 | 43 | 42 | 1.44 | 1.02 |
| 3b | 183 | 12.3–14.8 | 18 | 50 | 0.8 | 0.36* | 20 | 48 | 0.59+ | 0.42* |
| 7 | 127 | 34.3–39.7 | 32 | 26 | 2.73** | 1.23 | 29 | 29 | 1.4 | 1 |
| Overall | 1314 | NA | 217 | 467 | – | 0.46** | 296 | 388 | – | 0.76* |
The total number of QTL with reinforcing (n. reinforce) and antagonistic (n. antag.) effects are presented for each treatment and split by the position (if the trans-eQTL is in one of the three hotspots). Fisher’s exact tests for imbalance between antagonistic and reinforcing effects were conducted within each treatment with two NULL hypotheses: (1) the total bias towards reinforcing effects among all genes with cis- and trans-eQTL (v. background) and (2) a 1:1 ratio, expected under purely neutral evolution (v. neutral). Odds presented indicate the ratio of the observed bias toward reinforcement relative to the NULL ratio at each trans-eQTL location. Significance codes: **P-value<0.001, *P-value<0.05, +P-value<0.1.