Genome annotation remains remarkably incomplete. Even the genomes of prokaryotic “model organisms” are not comprehensively annotated - many translational events remain of dubious status and/or functionally uncharacterised, small RNAs are missing, and small proteins and overlapping genes are often also missed. In this talk I will present the findings of an integration of thousands of transcriptomes, protein structures, and genomes across the E. coli species, to discover new functional genetic elements and infer functions for some uncharacterised elements. The results presented are based on a pantranscriptome analysis of 15,000 RNAseq samples across 10 lineages of E. coli, supplemented with protein structures and analysis of selection. Genetic elements considered include non-coding RNAs, small proteins, and overlapping genes. This analysis extends prokaryotic pangenomics into pantranscriptomics as well as the “pan” study of protein structures and elements subject to natural selection across the species. As such, it outlines new directions for using pangenomes in aid of comprehensive bacterial genome annotation.