The GLP-1 family look alike on a specification sheet for a reason: they are variations on one natural sequence, made longer-lived by the same trick.

Why the length is what it is

Native GLP-1 is thirty amino acids. The analogues on this shelf run to thirty-one, thirty-nine or more because they add an extension at the C-terminus, and past about thirty residues every extra cycle costs real yield. Retatrutide at thirty-nine residues is not a little harder to make than a ten-residue peptide. It is a different order of difficulty.

The Aib substitution

Position 8 of the natural sequence is where dipeptidyl peptidase-4 cuts. Substituting aminoisobutyric acid there — a non-natural amino acid with no alpha hydrogen — removes the shape the enzyme is looking for. Semaglutide carries one; tirzepatide carries two.

The fatty diacid

The chain hanging off a lysine side chain is a C18 or C20 diacid, and it is there to bind albumin. A molecule travelling bound to albumin is not filtered out by the kidney at the rate a free one is, which is the whole of why these have the half-lives they do.

It is also a synthesis step that can fail, on a residue that has to be selectively deprotected while everything else stays protected. That step is a meaningful part of the price, and it is why two chains of identical length can differ severalfold in cost.

What that means for a certificate

The failure modes are specific. A missing diacid gives a molecule of nearly the right length and noticeably the wrong mass, which is exactly the case mass spectrometry catches and an HPLC purity figure alone does not — another reason identity and purity are separate questions.