Data presented in FBrf0225793 (Brown et al, 2014) and accompanying supplementary materials were used to identify potential new gene models and existing gene models in need of reassessment. Each model was assessed in the context of other available data, including stranded RNA-Seq expression data, transcription start site data, and tBLASTn analyses against other Drosophila species. All resulting new gene models are for non-protein-coding genes, primarily anti-sense RNA genes. Among the existing gene models that have been modified are 5 cases of genes previously annotated as non-coding that have been reannotated as protein-coding genes. The genes affected and the types of changes are enumerated in a spreadsheet available for download from the FlyBase ftp site (see link on this page). The specific source of the data for each, i.e, the published paper or in one of the supplementary materials, is also indicated.

Note that in Supplementary Table 7, "PhyloCSF scores for all ORFs between 20aa and 100aa in length," some ORFs within anti-sense genes (which often contain shadow coding regions) or within pseudogenes (which may contain intact regions of significant length) have PhyloCSF scores in the range of protein-coding exons. These have not been annotated as protein-coding regions in FlyBase.