Open Reading Frames
A cell reads a gene three bases at a time. Each triplet, a codon, stands for one amino acid or for "stop". An open reading frame (ORF) is an uninterrupted run of codons that starts at a start codon (ATG, which codes for methionine) and continues until the first in-frame stop codon (TAA, TAG or TGA).
Six reading frames
Where you start counting decides the triplets. For ATGGCCATTG…, frame +1 reads ATG GCC ATT G…, frame +2 reads TGG CCA TTG…, and frame +3 reads GGC CAT TG…. DNA has two strands and a gene can be on either one, so there are three more frames on the reverse complement (−1, −2, −3): six in all.
Finding an ORF by hand
Take ATGGCCATTGTAATGGGCCGCTGA. Start at the first ATG and read in threes: ATG GCC ATT GTA ATG GGC CGC TGA. That is Met-Ala-Ile-Val-Met-Gly-Arg, then TGA, a stop. The ORF is 24 bases long and codes for a 7-residue peptide. The second ATG inside it is in the same frame, so it is part of the same ORF, not a new one.
Is this ORF a real gene?
Stop codons are common. In random sequence with equal base frequencies, 3 of the 64 codons are stops, so a frame is interrupted on average every 64 ÷ 3 ≈ 21 codons. An ORF of 100 codons or more happens by chance in well under 1% of cases, which is why length is the first and best clue. Short ORFs are mostly noise.
Base composition changes the odds. The stop codons are AT-rich, so in a GC-rich genome they are rare and long, meaningless ORFs appear easily, often on the strand opposite a real gene. Treat a long ORF in GC-rich DNA with extra care.
What a simple ORF finder cannot see
Introns. Human and other eukaryotic genes are interrupted by introns that are cut out of the mRNA, so on genomic DNA the coding sequence is split into pieces and no single ORF covers it. Use the mRNA or cDNA sequence instead, or a gene-prediction tool.
Other start codons. Bacteria also start translation at GTG and sometimes TTG, and some genes lack a recognisable start in the part you pasted. Only ATG is searched here.
Other genetic codes. Translation uses the standard code. Mitochondria and a few organisms differ (in vertebrate mitochondria TGA codes for tryptophan, not stop).
How Aaron's Archive does it
| Item | What the tool does |
|---|---|
| Frames | All six, on both strands. Coordinates and frames are reported along the sequence you entered. |
| Start | ATG. Where several ATGs share one stop, the first (the longest ORF) is reported. |
| Minimum size | 30 bases including the stop (10 codons), so tiny accidental ORFs are not listed. |
| Output | An ORF map, a table sorted longest first with protein length and the start of the translation, and a six-frame translation view. |
Next steps
Check a promising ORF by translating it and running BLAST to see whether the protein resembles a known one. Open the protein properties to see what kind of protein it would make.
Common questions about ORFs
What is the difference between an ORF, a CDS and a gene?
An ORF is any open stretch you can find in sequence. A CDS (coding sequence) is the part of a gene that is actually translated, as annotated from evidence. A gene also includes regulatory and untranslated regions, and in eukaryotes introns.
How long does an ORF have to be to count?
There is no fixed rule. Many pipelines use 100 codons (300 bases) or more for a confident gene call, and shorter cut-offs when they also look for small proteins.
Why are there six reading frames?
There are three ways to group a sequence into triplets, and DNA has two strands that can each carry a gene, so 3 × 2 = 6.
Can I find ORFs online?
Use the ORF finder: it searches all six frames and lists each ORF with its translation.