Table 3. Performance of programs on the ENCODE IMR90 data: features (full-transcripts and introns) matching known and/or high-confidence novel annotations. GENCODE v.17 chromosome 1 annotation contains 15 493 transcripts and 33 202 introns. R = (recall) = Match/Annotations, P = (precision) = Match/Predicted.
| Program | Transcripts | Introns | ||||||
|---|---|---|---|---|---|---|---|---|
| Predicted | Match | R | P | Predicted | Match | R | P | |
| Cytosol | ||||||||
| CLASS2 | 3029 | 1053 | 0.068 | 0.348 | 12 662 | 12 557 (11 996) | 0.378 (0.361) | 0.968 (0.947) |
| CLASS2_F0.01 | 3836 | 1183 | 0.076 | 0.308 | 13 413 | 13 253 (12 327) | 0.399 (0.371) | 0.988 (0.919) |
| Cufflinks | 2508 | 621 | 0.040 | 0.248 | 10 420 | 10 372 (10 109) | 0.312 (0.304) | 0.995 (0.970) |
| Cufflinks_F0.01 | 3458 | 719 | 0.046 | 0.208 | 11 725 | 11 564 (10 779) | 0.348 (0.325) | 0.986 (0.919) |
| IsoCEM | 2479 | 722 | 0.047 | 0.291 | 11 483 | 11 297 (10 617) | 0.340 (0.320) | 0.984 (0.925) |
| Scripture | 14 621 | 971 | 0.063 | 0.066 | 13 820 | 12 751 (11 149) | 0.384 (0.336) | 0.923 (0.807) |
| iReckon | 4512 | 1730 | 0.112 | 0.383 | 11 724 | 11 477 (10 552) | 0.346 (0.318) | 0.979 (0.900) |
| Nucleus | ||||||||
| CLASS2 | 6084 | 992 | 0.064 | 0.163 | 16 391 | 15 765 (12 862) | 0.475 (0.418) | 0.962 (0.846) |
| CLASS2_F0.01 | 10 216 | 1141 | 0.074 | 0.112 | 18 610 | 17 699 (14 539) | 0.532 (0.438) | 0.950 (0.781) |
| Cufflinks | 2714 | 561 | 0.036 | 0.207 | 11 255 | 11 079 (10 576) | 0.334 (0.319) | 0.984 (0.940) |
| Cufflinks_F0.01 | 6085 | 789 | 0.051 | 0.130 | 16 884 | 16 064 (13 568) | 0.484 (0.409) | 0.951 (0.804) |
| IsoCEM | 2236 | 277 | 0.018 | 0.124 | 9604 | 8737 (7576) | 0.263 (0.228) | 0.910 (0.789) |
| Scripture | 45 247 | 764 | 0.049 | 0.017 | 18 048 | 13 910 (10 188) | 0.419 (0.307) | 0.771 (0.564) |
| iReckon | 5769 | 1539 | 0.099 | 0.267 | 10 162 | 9474 (8232) | 0.285 (0.248) | 0.932 (0.810) |