This track shows 22,567 upstream open reading frames (uORFs) in the 5' untranslated regions (5' UTRs) of human protein-coding genes, compiled as part of the 5ULTRA pipeline for annotating 5' UTR variants. The uORFs are defined on MANE Select transcripts, which provide a single well-supported, clinically relevant transcript per gene matched between Ensembl/GENCODE and RefSeq. Only ATG-initiated uORFs are included.
uORFs are short open reading frames found upstream of the main protein-coding sequence. When a ribosome scans from the 5' cap, it may translate a uORF before reaching the main start codon, which often reduces production of the downstream protein. Genetic variants that create, disrupt, or alter uORFs can therefore change protein output and contribute to disease.
Each uORF is classified into one of three types based on the position of its stop codon relative to the main CDS start:
Items are colored by Kozak context strength, using the same color scheme as all other subtracks in this collection:
| Strong: both position −3 (A/G) and position +4 (G) match. | |
| Moderate: only one of position −3 (A/G) or position +4 (G) matches. | |
| Weak: neither position −3 (A/G) nor position +4 (G) matches. | |
| NA: Kozak context could not be scored (465 uORFs). |
uORFs can span multiple exons. Each uORF borrows its exon/intron structure from an overlapping MANE Select transcript. If no MANE transcript has introns that match the uORF's endpoints, the GENCODE comprehensive annotation is used instead. The source transcript is recorded in the intronsSource field (none if no matching transcript was found).
The track can be filtered by:
Mouseover shows the gene symbol, uORF type, rank within the gene, Kozak strength, and the donor transcript used for intron structure.
The data can be explored interactively in table format with the Table Browser or the Data Integrator and exported from there to spreadsheet or tab-separated tables. From scripts, the data can be accessed through our API, track=fiveUltraUorfs.
For automated download and analysis, the genome annotation is stored in a bigBed file that can be downloaded from our download server. The file for this track is called fiveUltraUorfs.bb. Individual regions or the whole genome annotation can be obtained using our tool bigBedToBed, which can be compiled from the source code or downloaded as a precompiled binary for your system. Instructions for downloading source code and binaries can be found here. The tool can also be used to obtain features within a given range, e.g.
bigBedToBed http://hgdownload.soe.ucsc.edu/gbdb/hg38/ncOrfs/fiveUltraUorfs/fiveUltraUorfs.bb -chrom=chr21 -start=0 -end=100000000 stdoutThe original uORF reference set is distributed with the 5ULTRA software package and can be obtained by installing the package and running 5ULTRA-download-data.
Chaldebas et al. compiled a reference set of uORFs from three sources: the Ribo-uORF database (501,554 translated uORFs supported by 1,495 ribosome profiling datasets), the uORFdb database of computationally predicted uORFs, and the sorfs.org database. Only ATG-initiated uORFs were retained. These were mapped to the 5' UTRs of 18,775 MANE Select protein-coding transcripts from GENCODE v45 basic annotation, yielding 22,567 uORFs. Of these, 8,067 (35.7%) have direct ribosome profiling support. Each uORF was classified as Non-Overlapping, Overlapping, or N-terminal extension based on the relationship between its stop codon and the main CDS start. See Chaldebas et al. 2026 for full details.
The uORF reference BED file was obtained from the 5ULTRA GitHub repository (installed via 5ULTRA-download-data; file uORFs.MANE.hg38.bed). Colors were remapped to distinguish Kozak strength (see Display Conventions). To recover exon/intron structure for uORFs that span introns, each feature was projected onto an overlapping MANE Select transcript using the addIntrons.py script in the kent source tree; a GENCODE v49 comprehensive bigBed was used as fallback when no MANE candidate could provide intact boundaries at the uORF endpoints. Of 22,567 uORFs, 3,861 received multi-block exon structure from MANE, 9 from GENCODE fallback, and 45 overlapped a GENCODE transcript but had no introns within the uORF range; 18,652 remain as single-exon features whose genomic span contains no intron from any known host. Build steps are recorded in the makedoc; the processing scripts are at src/hg/makeDb/scripts/ncOrfs/.
Thanks to Matthieu Chaldebas and the 5ULTRA team for making the uORF reference data publicly available as part of the 5ULTRA package.
Chaldebas M, Ponsin K, Bohlen J, Conil C, Mourelatos H, Stenson PD, Cooper DN, Abel L, Casanova JL, Cobat A et al. Genome-wide detection of human 5' UTR variants that impact protein translation. Am J Hum Genet. 2026 Apr 2;113(4):809-827. PMID: 41881026; PMC: PMC13087467