Table 1.
Marker classes identified by random GFP∷cDNA screening
| Localization phenotype | Number isolated | Correct frame |
|---|---|---|
| Torus | 43 | 1 /14 |
| Nuclear exclusion | 20 | 8 /10 |
| Q-balls | 8 | 0 /2 |
| Nuclear localization | 7 | 2 /3 |
| Streaming dots | 7 | 1 /3 |
| Cell surface | 5 | 5 /5 |
| Nucleolus | 6 | 2 /4 |
| Bright nuclei | 5 | 0 /2 |
| Blobs | 3 | 1 /2 |
| Vacuolar membrane | 4 | 3 /3 |
| ER membrane | 2 | 1 /1 |
| Cell contact junctions | 3 | 0 /1 |
| Tiny bubbles | 3 | 2 /2 |
| Chromosomes | 1 | 1 /1 |
| Regulated nuclear exclusion | 1 | 1 /1 |
| Wound induced granulation | 1 | 1 /1 |
| Darts | 2 | 0 /1 |
| Total | 120 | 29 /56 |
Each of 120 transgenic lines was classified into one of 17 types (left column) based on the appearance of GFP localization by confocal microscopy. The number of lines isolated in each phenotypic group is shown in the middle column. The frequency with which fusion proteins of a given phenotypic group were in the same translational reading frame as their native proteins is shown in the right column. Thus, for instance, the correct reading frame could be inferred for 14 of the fusion proteins in the “Torus” class, but only 1 of these was in frame with GFP. These estimates were inferred by sequence analysis of the subset of single insert transgenic plants containing markers with homology to previously characterized genes (see text for description).