Table 2. Sequence validation rates of predictions, FGENESH vs. GENSCAN

Gene prediction set	Gene prediction program	Total no. of predictions	No. of predictions tested*	No. sequence-validated (%)^†
sjc set	GENSCAN	196	161	60 (37.3%)
sjc set	FGENESH	11	10	4 (40.0%)
Heidelberg set^‡	FGENESH	1,266	160	18 (11.3%)
Homol-2 set	GENSCAN	333	204	27 (13.2%)
Homol-2 set	FGENESH	6	5	2 (40.0%)
Homol-0 set^§	GENSCAN	9,463	129	7 (5.4%)
Homol-0 set^§	FGENESH	127	75	5 (6.7%)
Total predictions		11,402	744	123 (16.5%)
Control set		159	159	154 (96.9%)

*Gene predictions from each prediction set (sjc set, comprised of GENSCAN or FGENESH predictions with an intron conserved in D. pseudoobscura; Heidelberg set, FGENESH predictions reported verified by transcription profiling (1); homol-2 set, with homology to D. pseudoobscura in two exons; homol-0 set, with homology in zero to one exon; and release 3.1 controls) were tested by sequencing of RT-PCR products.

^{†Gene predictions were considered validated if the aligned sequence of the PCR product was consistent with a spliced gene model in the region of the prediction.}

‡Only 1,266 multiexon predictions from the 2,636 predictions described in the Hild et al. (1) study were considered for analysis, and, of these, we tested only the 160 with the highest priority scores that did not overlap any GENSCAN or FGENESH predictions tested in the other sets.

^{§The homol-0 set was selected to be representative of the full range of priority scores.}