Lagarde, J. et al. Nat. Genet. 49, 1731–1740 (2017).

To begin to understand the function of a cell's large repertoire of long noncoding RNAs (lncRNAs), one must start with their detailed annotation. Current resources are either based on automated annotation and are therefore large but have incomplete transcript structures, such as the 101,700 genes in NONCODE, or they are manually annotated and therefore accurate but small, such as ENCODE's 15,767 lncRNA genes. Lagarde et al. seek to increase the number of lncRNAs in the ENCODE resource by a combination of targeted RNA capture and PacBio single-molecule real-time sequencing. They profile full-length lncRNAs in human and mouse tissues and present new transcript models for over 3,000 human and over 500 mouse genes.