Genome‐Wide Maps of the Human Transcriptome Reveal an Interleaved Organization and Novel Short and Long Classes of RNAs

Genomic regions not coding for proteins nor involved in cis or trans‐acting regulatory activities are at best viewed as sites for evolution‐mediated experimentation of novel functional domains and at worst, “junk”. However, recent unbiased empirical genomic wide have revealed that there are large portions of the non‐protein coding are transcribed in a regulated fashion in several organisms including humans. A recurrent question derived from these recent findings centers upon the biological functions of the observed unannotated transcripts. Considerable portions of the unannotated transcription observed in these species serve as unreported parts of protein coding transcripts and as precursors to the generation of short RNAs. As part of the Encyclopedia of DNA Elements (ENCODE) project, studies reveal that more than half of the well characterized protein coding loci present in ENCODE designated human genome regions (∼400) make use of tissue specific alternative, unanntoated 5′ transcriptional start sites (TSS) which for as much as 30% of the extended transcripts, are more than 200,000 nucleotides (nt) upstream of the annotated portions of the transcripts (108,000 nt on average). We have made similar observations in the organization and regulation of transcription during the first 24 hours of development of Drosophila melanogaster. In addition, often sharing the same genome sequences as protein coding transcripts are distinct l nuclear transcripts which serve as primary transcripts for the formation of small RNAs ranging in size from∼20–200 nt. Recently, two new classes of short RNAs have been identified from our analyses of the sites of transcription across the human genome. These data support a new model for genome organization which is not co‐linear but rather lattice‐like and in part focused on the production of short RNAs.