Table 2.
Results of an evaluation comparing exonerate (using the methods presented here), with sim4 and est2genome. A subset of the ensembl database was used, containing 1,827 genes, 9,917 exons and 2,387,971 bases. Times are given for alignment of all cDNAs to all contigs. The CPU time for est2genome was estimated from a smaller number of alignments. The accuracies are comparable, but the times are much faster for the heuristic methods.
Method | Base Accuracy | Exon Accuracy | CPU Time (seconds) | x-fold |
est2genome | 99.99% (2,387,648) | 99.90% (9,907) | ≅ 7.2 years (226,637,523) | 1 |
sim4 | 98.97% (2,363,438) | 98.12% (9,731) | ≅ 60 hours (216,956) | 1045 |
exonerate | 99.88% (2,384,988) | 99.98% (9,916) | ≅ 46 minutes (2,577) | 87946 |