Open Reading Frame Calculator

Calculate ORF properties, reading frames, and expected coding sequence lengths.

Standard: 3 (UAA, UAG, UGA)
Please check your inputs and try again.

Why Long ORFs Are Significant

In random sequence, three of the 64 codons are stop codons, so a stop appears on average every 64/3 ≈ 21 codons. Long open reading frames therefore occur by chance far less often than intuition suggests.

P(ORF ≥ n codons) ≈ (1 − 3/64)n
ORF lengthProbability by chanceAssessment
21 codons~36%Expected average
50 codons~9.1%Unremarkable
100 codons~0.82%Probably real
> 300 codons< 10−6Almost certainly a gene

This is why a 100-codon threshold is the common default for ORF finding. Below it, chance ORFs outnumber genuine ones; above it, the reverse holds sharply.

Why GC Content Changes the Threshold

All three stop codons — TAA, TAG, TGA — are AT-rich. In a GC-rich genome they arise less often by chance, so random ORFs are longer and the significance threshold must rise.

Genome GC%Effect on chance ORFsExample organism
30%Shorter — stops frequentPlasmodium
50%BaselineE. coli
70%Longer — stops rareStreptomyces

Applying an E. coli threshold to a Streptomyces genome would produce many false positive genes. Modern gene finders build organism-specific models rather than using a universal cutoff.

Six Frames, and What ORF Length Cannot Tell You

Double-stranded DNA has six reading frames — three on each strand. All must be searched, and overlapping ORFs on opposite strands are common in compact genomes such as viruses.

Length alone is a weak criterion in eukaryotes, where introns interrupt coding sequence so genomic ORFs are short even for long genes. Ribosome profiling has also revealed many functional short ORFs that length-based finders discard, so the 100-codon convention excludes real biology as well as noise.

Worked Examples

Example 1: 900 nt sequence (300 codons)
Expected ORF by chance: 64/3×3=64 nt
Result: 300 codon ORF is highly significant
Likely a real gene
Example 2: 50% GC genome, stops=3/64
Avg ORF by chance: 64 nt
Result: ORFs >200nt are real
Prokaryotes: minimal UTR, dense coding
Example 3: Effect of GC content
Same 300-codon ORF in a 30% versus 70% GC genome
Result: Far more significant in the AT-rich genome
Stop codons are AT-rich and therefore common in AT-rich sequence. The same ORF length is much easier to achieve by chance at high GC.
Example 4: Overlapping genes
Compact viral genome, ORFs on both strands
Result: Six-frame search essential
Overlapping and antisense genes are common where genome size is constrained. Searching three frames would miss a substantial fraction.

Common Mistakes

⚠️
Using one ORF threshold across all genomes

Stop codons are AT-rich, so GC-rich genomes produce longer chance ORFs. An E. coli cutoff generates false positives in Streptomyces.

⚠️
Searching only three reading frames

Double-stranded DNA has six. Genes on the reverse strand are missed entirely if only the forward frames are examined.

⚠️
Applying ORF length directly to eukaryotic genomic DNA

Introns break the reading frame, so genomic ORFs are short even for long genes. Use cDNA, or a gene finder that models splicing.

⚠️
Assuming short ORFs are never real

Ribosome profiling has identified many functional peptides under 100 codons. The conventional threshold trades sensitivity for specificity.

Frequently Asked Questions

How many reading frames?
6 total: 3 forward (+1, +2, +3 starting at position 1, 2, 3) and 3 reverse complement (-1, -2, -3). All must be searched. Prokaryotes: ORFs in all 6 frames possible (no introns). Eukaryotes: ORF spans exons, complicated by splicing.
ORF finding tools?
NCBI ORF Finder, getorf (EMBOSS), Augustus, GeneMark. Parameters: minimum ORF length (usually 100 codons), codon table, input circular/linear. Compare with BLAST hits to confirm biological relevance.
How long must an ORF be to be significant?
Around 100 codons is the usual default. Chance ORFs average about 21 codons, and the probability of reaching 100 by chance is under 1%.
Why does GC content matter?
All three stop codons are AT-rich, so they occur less often in GC-rich sequence. Random ORFs are longer, requiring a higher significance threshold.
How many reading frames are there?
Six — three on each strand of double-stranded DNA. All must be searched, since genes occur on both strands.
Why is ORF length unreliable in eukaryotes?
Introns interrupt the coding sequence, so genomic ORFs are short even for long genes. Splice-aware gene finders or cDNA are needed instead.