Previously unannotated ORFs encode functional proteins involved in cell biology

Isabelle Fournier, Michel Salzet

Trends Open

https://doi.org/10.1016/j.treopn.2026.04.002

Abstract

Proteins encoded by previously unannotated open reading frames (ORFs) are increasingly recognized as common products of eukaryotic genomes, challenging conventional gene annotation. This question is especially timely because ribosome profiling, long-read transcriptomics, and proteogenomics now reveal widespread translation beyond annotated coding regions. The central challenge is to discriminate functional proteins from pervasive, potentially nonproductive translation. Recent progress addresses this by combining transcript-aware ORF mapping, evidence of translation, protein-level detection, and genetic validation. These approaches also sharpen the distinction between transcript architecture and the translation mechanisms that generate diverse protein outputs. Overall, this review argues that gene output is inherently multiproduct and that a rigorous cell-biological framework is required to systematically identify and study functional proteins encoded by previously unannotated ORFs.