DNA is written in a four-letter alphabet. So is base 4. Turn the digits of π, φ, e, √2 and √3 into A, T, G and C, and they contain sequences that a cell could read as proteins. Is that meaningful, or just chance? This site lets you check.
Everything below is computed live in your browser, using the method from the paper.
The first 2,000 decimal digits after the point: π → 1415926535…
Every digit 0–9 becomes two base-4 digits. 7 → 13, 9 → 21.
0→A 1→T 2→G 3→C, then keep the first 1,500 bases.
Find every ATG start codon, translate until a stop codon, and keep proteins of 5+ amino acids.
Pick a constant to see its DNA, its open reading frames (protein-coding stretches) and the proteins they translate to. Click a protein to find it in the sequence.
| Frame | Position | Length | Protein |
|---|
The key question the paper raises: do constants produce more or longer proteins than random digits would? Here we run the exact same pipeline on thousands of random digit strings and see where each constant lands.
Paste any string of decimal digits: another constant, a phone number, a date. Non-digit characters are ignored.
Mathematical Constants as Generative DNA Seeds: A Computational Hypothesis for Irrational Numbers as Protein Sequence Generators.
Mohcin Bajja, Independent Researcher, Khémissat, Morocco. Preprint v1.0, April 2026.
When mapped to DNA through base 4, all five constants contain open reading frames and translatable proteins, with GC content around 29–32% and longest proteins of 40–73 amino acids. √3 gives the longest (73 aa); e gives the most diverse.
No biological origin, no design. The hypothesis is computational and falsifiable: do constant-derived sequences differ from random sequences of the same composition? The paper asks the community to test it with random baselines, BLAST and protein folding tools such as AlphaFold2 or ESMFold.
Questions, replications or collaboration: mohcin15000@gmail.com