What are peptides?

An amino acid is a small molecule with two halves. One half is identical in every amino acid — a standard connector, the same every time. The other half, called the side chain, is what makes each one different: greasy or water-loving, big or tiny, positively or negatively charged.

Life uses just 20 standard amino acids. That's the entire alphabet. Every protein in every living thing — your eyes, an oak tree, a virus — is spelled out of these same 20 letters.

Click amino acids together into a long chain and something remarkable happens. The chain doesn't stay limp. Its side chains push and pull on each other — water-avoiding parts hide inside, charged parts grab their opposites — and the whole strand collapses into one specific, repeatable 3-D shape.

That shape is the function. A protein folded into a pocket can grip a molecule. Folded into a channel, it lets things through a wall. Folded into a claw, it copies DNA. Proteins are the machines, scaffolding, and workforce of every cell — and a protein is nothing more than a chain of our 20 letters, folded.

A peptide is a short chain of amino acids — the same letters, the same kind of bond, just fewer of them. A peptide is a word; a protein is a paragraph. Both are written in the same alphabet.

Here's the honest part most explanations skip: there is no strict, official line where "peptide" ends and "protein" begins. The convention is roughly 50 amino acids — below that, people usually say peptide; above it, protein — but it's a soft boundary, not a law. Drag the slider and watch one become the other.

A body is trillions of cells that can't see or talk to each other. They need a way to coordinate — to say "store this sugar," "raise the alarm," "grow now," "you can stop." Peptides are one of biology's favorite ways to send those messages.

Their shortness is the point. A peptide is big enough to carry a specific instruction but small enough to be made quickly, sent fast, and broken down soon after — so the message is heard and then it's gone, instead of lingering forever. Many of the hormones you've heard of — insulin, oxytocin — are peptides doing exactly this.

The information isn't written in words — it's written in shape. One cell builds a peptide with a very particular form and releases it. It drifts through the bloodstream, passing nearly every cell in the body. To almost all of them, it's invisible. Only a cell built to recognize that exact shape will respond.

The cell's reader is called a receptor: a protein that sits in the cell's outer wall with part of it facing out, shaped to fit one specific peptide and nothing else. When the matching peptide settles into it — like a key into a lock — the receptor changes shape.

That shape-change on the outside flips a switch on the inside, kicking off a chain reaction that reaches into the cell and changes its behavior: switch on a gene, release a hormone, start dividing, or stand down. Try it — send each peptide and see which one the receptor accepts.

Every peptide so far has been a go signal: it fits a receptor and switches something on. But fitting and activating are two different things. A peptide can be built to slot into a receptor's pocket perfectly and still not flip the switch — and when it does that, it stops being a message and becomes a blocker.

Sitting in the pocket, it leaves no room for the real signaling molecule. The natural message arrives, finds its spot already occupied, and goes unread. Chemists call this competitive inhibition — two molecules competing for the same site, with the outcome tipping toward whichever is present in greater force. The same idea works on enzymes: park a peptide where an enzyme grips its target, and the enzyme can't do its job.

That choice — activate or block — is much of what makes designed peptide binders useful. A great many are built not to speak but to get in the way on purpose: to quiet an overactive receptor, occupy an enzyme's active site, or shield a target from the molecule that would otherwise switch it on. The same precision that lets a peptide deliver an instruction is exactly what lets it intercept one.

If a peptide is a message your body already understands, then a peptide medicine is a message we write on purpose. This isn't speculative — it's been saving lives for a century.

Insulin, the peptide that tells cells to take up sugar, became the first peptide medicine in the 1920s. A century later, GLP-1 medicines — redesigned copies of a gut hormone — are reshaping how we treat diabetes and obesity. Trace the arc:

Finding a new peptide medicine means answering one hard question: out of an almost unimaginable number of possible sequences, which rare few will do a useful job in the body safely? A chain just ten letters long already has more possible sequences than there are stars in the galaxy.

So discovery is a loop. Pick a target, design candidate sequences, build the promising ones — assembling each peptide one amino acid at a time — then test them against real biology, and let every result teach the next round. Step through it:

The bottleneck has always been the size of the search. You cannot build and test billions of peptides in a lab — there isn't enough time, money, or material in the world. This is where modern AI changes the economics.

Machine-learning models, trained on decades of biological data, can predict what shape a sequence will fold into and how tightly it might grip a target — for millions of candidates at once, before a single one is made. They turn an impossible search into a short list worth testing.

Because peptides speak the body's own language, they can be aimed with a precision that blunter drugs can't match. A few of the areas where they already matter — or soon could:

For most of history, medicine found peptides by luck — stumbling on insulin, isolating a hormone, copying what nature already made. The space of possible peptides, the ones evolution never got around to trying, is almost entirely unexplored. Somewhere in it are treatments that don't exist yet.

The work now is to search that space deliberately: to combine the precision of the body's own language with computation that can explore millions of possibilities and lab science that can tell us which are real. That search is what Protean Labs is built to do.