Into the glow

How it was made

1. Real data. Atomic coordinates from PDB entry 1EMA: X-ray crystallography at 1.9 Å, Ormö et al. 1996, released under CC0.
2. A pipeline, not a drawing. A Python script (gemmi, NumPy) reads the file, keeps the 1,771 atoms and the 11 strands, aligns the barrel and writes 42 KB of data.
3. One file. That data is embedded in a single HTML page that renders everything with Three.js. No server, no framework.
4. Every fact traced. Each fact on screen traces back to a primary source: the structure file, its publication, the PDB data policy or the Nobel Prize documentation. One detail stayed flagged until the primary source was found.
5. Two reviews. A protein scientist checked the science. Javier Pino checked it as training content.
6. Tested like software. Screenshots of every stop on desktop, mobile, light and dark, taken with Playwright in a headless browser.
7. Built with Claude Opus 5.5 (Anthropic), directed and verified by Javier Pino. Code and data: github.com/jpinob/into-the-glow.