115-Million-Year-Old Blood and the Case for Walking Inside Molecules

My grandmother kept a shoebox of things that shouldn't have survived: a dried corsage from 1962, a train ticket from a town that no longer exists, a photograph so faded it looks like weather. She'd open it on Sunday afternoons and say, look how much of it is still here.

I thought about that shoebox this week, when the news broke that scientists had chemically confirmed traces of ancient blood inside 115-million-year-old dinosaur bones. Not a guess. Not a stain that resembled something. Chemistry.

A hundred and fifteen million years. Through ice, through heat, through being buried and crushed and forgotten. And some small architecture of that animal is still holding its shape.

I'm a biochemist. I teach second-year structural biology, and I have spent eleven years telling students that a protein is a machine with a shape. That's the sentence. That's the whole first lecture. And every year, at least one student nods politely and then fails the docking question on the exam, because they've never actually seen a pocket. They've seen a picture of one.

Here's the part I don't usually admit out loud: I hadn't really seen one either. Not the way I wanted to.

Last spring I sat in my office at 11:40 at night, the kind of quiet where you can hear the building's ventilation, rotating a structure on my monitor with a trackpad. I had a ligand sitting in a binding site and I was trying to understand why it fit — why there and not forty angstroms over, where the surface looked just as plausible. I zoomed. I rotated. I changed the representation six times. And I kept losing the third dimension somewhere between my eyes and the screen. Flat representations of a three-dimensional problem. I closed the laptop with that specific tiredness that isn't about sleep.

A colleague handed me a headset a few weeks later and said, just try the hemoglobin tour.

I want to be honest about what happened next, because it sounds dramatic and it wasn't. It was just... obvious. I was standing inside hemoglobin at my own scale. Oxygen came in and docked at the heme, and I turned my head and watched the pocket from the side, and I understood — not intellectually, not from a figure — that this is a space. A room with walls and a floor and a doorway, and the ligand walks in through the doorway.

That's it. That was the whole shift.

I went home and stayed up past one, going through the other tours like someone flipping through a book they can't put down. Standing in the middle of glycolysis, watching the glucose come apart step by step. Watching aspirin settle into a COX enzyme and block it. CRISPR-Cas9 opening DNA. Twelve narrated tours across six pillars — Biotechnology Basics, Foundations of Life, Pathways & Interactions, Medicines in Action, Genetic Technologies, Green Biotech — all running on real structures from the Protein Data Bank. My eleven-year-old sentence about machines with shapes finally had a room to live in.

The part that changed my research life came later, in the professional viewer. I typed a gene name into the search bar and it cascaded across RCSB PDB, UniProtKB, AlphaFold DB, SWISS-MODEL and PDBe — resolving synonyms I would have chased manually across five tabs. The structure materialized around me. I cycled the Mol* presets with one tap — cartoon, ball-and-stick, spacefill CPK, molecular surface, B-factor heat, tube trace — until the thing looked the way it looked in my head.

Then I ran AI pocket prediction on a protein I'd been stuck on. A druggable site lit up. I walked around it. I grabbed it with my hands and scaled it down and up again. I used the Ligand Inspector to trace the hydrogen bonding network, and I picked two atoms with the trigger just to watch the distance report in angstroms on the wall in front of me. Then I superimposed a homolog with the Foldseek pipeline and asked a question I'd been circling for months: does the function conservation hold here, or not?

It didn't answer the question for me. That's the point. It made the question visible, which is a different and rarer gift. I sat down the next morning and rewrote my docking hypothesis in about forty minutes. Three weeks of circular trackpad work, resolved by standing up.

I still open the shoebox when I visit home. Still faded, still improbable, still there.

That's what the dinosaur blood story really says to me. Molecular information persists far longer than we assume — and our job, as people who teach it and study it, is to make it legible at the scale where intuition actually happens. For most of a century, that scale has been the flat rectangle in front of us. It doesn't have to be.

If you've ever zoomed until your eyes hurt and still not quite felt the pocket, I'd gently suggest you stop zooming. Stand up. Walk around the thing. It's on Meta Quest, and you can start with a tour you already know by heart.

NexoPeek is educational and research-visualization software — not medical advice, diagnosis, or treatment. Analysis aids such as pocket prediction, alignment, and ligand inspection support, never replace, validated scientific pipelines and professional judgment.

Step Inside the Molecule. Explore the Unseen World: https://www.meta.com/experiences/nexopeek/1453797551140347/