GASPing for genes

In the Genome Annotation Assessment Project (GASP), 12 groups tested their ability to identify genes in 3 million base pairs of Drosophila DNA. The results are reported in eight papers in the April issue of Genome Research. The majority of the groups correctly identified over 95% of the coding nucleotides. Less successful were predictions of intron/exon boundaries, which were correct for just over 40% of the genes, and predictions of promoter locations, which suffered from a high rate of false positives.