This article by Ian Korf introduces SNAP, a high-performance ab initio gene-finding program designed to be adaptable to various genomes. The author demonstrates that SNAP outperforms other gene finders like Genscan, HMMGene, Genefinder, and Augustus in multiple genomes, including Arabidopsis thaliana, Caenorhabditis elegans, Drosophila melanogaster, and Oryza sativa. However, when applied to novel genomes without an appropriate gene finder, foreign gene finders can produce highly inaccurate results, as their parameters are often not species-specific. Korf proposes a bootstrapping procedure to estimate species-specific parameters from foreign gene finder predictions, which can significantly improve the accuracy of gene prediction in novel genomes. The study highlights the importance of having dedicated gene finders for each genome due to the sensitivity of gene prediction to species-specific parameters.This article by Ian Korf introduces SNAP, a high-performance ab initio gene-finding program designed to be adaptable to various genomes. The author demonstrates that SNAP outperforms other gene finders like Genscan, HMMGene, Genefinder, and Augustus in multiple genomes, including Arabidopsis thaliana, Caenorhabditis elegans, Drosophila melanogaster, and Oryza sativa. However, when applied to novel genomes without an appropriate gene finder, foreign gene finders can produce highly inaccurate results, as their parameters are often not species-specific. Korf proposes a bootstrapping procedure to estimate species-specific parameters from foreign gene finder predictions, which can significantly improve the accuracy of gene prediction in novel genomes. The study highlights the importance of having dedicated gene finders for each genome due to the sensitivity of gene prediction to species-specific parameters.