Peptides in Metabolic Research: A Reading Map of the Preclinical Literature

Written by Marion Kessler · Reviewed by Douglas ReyesIndependent compilation of public literature; no institutional affiliation.

Corrections welcome. If a summary here misreads a paper, a note with the citation gets it fixed in the next pass. The notebook answers questions about its own sources only - never about sourcing, dosing or anything clinical.

The search phrase that brings most readers here is short and certain about itself: "What is the strongest peptide for fat loss?" I would rather answer it politely than pretend I can answer it as asked, so let me say at the top what this page is and what it is not. It is a reading map of laboratory research material, compiled by one person working through the metabolic peptide literature. It is not a statement about outcomes in people, because the record I can reach does not support one and I am not going to manufacture the sentence the query seems to want.

What I can describe honestly is the shape of the literature itself: which compounds recur in it, what receptor profiles they are described with, how far their study record goes, and how densely the papers cluster around each one. Those are four factual things about a published body of work. "Strongest" is not among them. In this field a potency figure is bound to a receptor, a cell background, a readout and a paper, and when those four do not match between two studies, the comparison quietly stops being a comparison.

So I keep the original wording visible and answer a different question underneath it. Below I work through the incretin-family compounds, the fragment literature, and the growth hormone axis peptides that tend to sit beside them, then place my own notes on ps r3, ps t2 and ps cg where they actually sit in that map. Everything here concerns laboratory research material only. The structural background is in my what is ps r3 peptide pillar page, and the handling side is in ps r3 peptide reconstitution.

Why I rewrite the question before I answer it

I want to spend real time on the rewrite, because it is the part of this page that does the work. When someone types "What is the strongest peptide for fat loss?" the intent is usually to get one name and one ranking. The literature does not produce that object. It produces receptor pharmacology tables, cell model readouts and study reports with inclusion criteria attached, and assembling those into a league table requires a chain of judgement calls that the papers themselves decline to make.

The first judgement call is which receptor counts. A compound described at two incretin receptors is not automatically ahead of one described at a single receptor, because the second receptor may contribute to the measured readout, may be neutral with respect to it, or may add a signalling limb that was never the point of the experiment. The second call is the preparation. Affinity obtained in membranes expressing a receptor at one density does not transfer cleanly to a cell line expressing it at another, and neither transfers to tissue.

The third call is which readout is allowed to stand in for strength, and this is where informal rankings break. Cyclic AMP accumulation, arrestin recruitment, receptor internalisation and downstream kinase phosphorylation each describe a different limb of the same receptor's behaviour, and one ligand can order differently on each. I go through the multi-agonist version of that problem in glp-3 rt peptide mechanism, where adding receptors makes the comparison sharper rather than easier.

The compounds that recur in metabolic research, and the columns I read them by

tirzepatide and retatrutide are trade names associated with eli lilly, and semaglutide is a trade name associated with novo nordisk. I cite all three for receptor-profile comparison only, and I imply no endorsement, affiliation or commercial relationship of any kind. The same applies to aod-9604 and cjc-1295, which I use purely as literature labels for the fragments and axis peptides that appear in this body of work.

I should say plainly where the compounds in the table below come from. They are not a list I assembled to answer "What is the strongest peptide for fat loss?" They are the names that recur in the metabolic peptide literature often enough that any reading of this field has to account for them, and I keep them in one table because that is the only way I can see the differences in record quality at a glance.

The incretin group is where the literature is densest, and it is also where the vocabulary is most stable. glp-1 class peptides are described as agonists at a class B G protein coupled receptor, and the papers around them report competition binding, second messenger accumulation and trafficking in a way I can follow without guessing. semaglutide is described in the same receptor terms with a formulation story attached, which is why I read it when I want to see how exposure changes without the receptor changing.

tirzepatide is described in the literature as a dual agonist at the gip and glp-1 receptors, and retatrutide as a triple agonist that adds the glucagon receptor. Both are useful to me for the same narrow reason: they are the clearest published examples of what adding a second or third receptor does to a signalling profile. That is a pharmacological question, not a ranking question, and it is the only reason they appear in a notebook about a different compound.

The fragment and axis material is thinner and I read it differently. aod-9604 is described as a c-terminal fragment related to growth hormone, with a receptor assignment that the papers I can reach do not settle. cjc-1295 is described as a growth hormone releasing hormone analogue, which puts its readouts at the level of an axis rather than a single receptor. Both show how little published work is needed before a compound becomes a confident search term.

How I read the compounds that recur in metabolic peptide literature, and what each column actually reports
Compound	Receptor profile as described	Where the study record stands	What I read it for
glp-1 class peptides	Described as agonists at the glp-1 receptor, a class B G protein coupled receptor	The densest literature of the group; many independent groups publishing	The reference standard I compare other profiles against
semaglutide	Described as a long-acting glp-1 receptor agonist	Extensive record including late-phase study reports in public registries	How a formulation change alters exposure without changing the receptor
tirzepatide	Described as a dual agonist at the gip and glp-1 receptors	Late-phase study reports alongside a large preclinical pharmacology record	What a second incretin receptor adds to the signalling profile
retatrutide	Described as a triple agonist at the glp-1, gip and glucagon receptors	Mid-phase study reports; preclinical literature still growing	Whether a third receptor widens the readout or complicates it
aod-9604	Described as a growth hormone related fragment; receptor assignment unsettled	Sparse; mostly older preclinical and in vitro reports	How thin a record can be and still anchor a search term
cjc-1295	Described as a growth hormone releasing hormone analogue acting at axis level	Limited and uneven; largely assay-level and preclinical work	How an axis readout differs from a receptor-level readout
ps r3	Receptor profile not verified in any primary paper I can reach	No peer-reviewed primary characterisation I can cite	A naming problem before it is a pharmacology problem
ps t2 and ps cg	Not established in the material I can reach	No primary characterisation I can cite for either	How sibling codes inherit claims from one another

What a receptor profile does and does not tell you

A receptor profile is the most quotable thing about any of these compounds and the easiest to over-read. It tells me which receptors were expressed in a preparation and which ligand was shown to engage them under stated conditions. It does not tell me that the list is complete, because a paper reports the receptors it tested, and a profile described as selective is only ever selective with respect to the panel that was run.

It also does not tell me about the limbs that were not measured. Two ligands can occupy the same receptor and produce different second messenger and trafficking behaviour, which is the basis of the way biased agonism is discussed in these papers. When a page says a compound is ahead of another because it acts at more receptors, what I want to know is which limb was measured, in which background, and whether anyone repeated it.

The honest version of the comparison is therefore a matrix rather than a ladder, and matrices do not survive being turned into a headline. None of this is a protocol for administration in humans or animals; it is how the material is described in the methods sections I read, and I keep it at that level on purpose. I write the assay side of this out more slowly in ps r3 peptide research applications.

A described receptor profile is bounded by the panel that was actually tested in that paper.
Occupancy and activation are different quantities and can order differently for the same ligand.
Second messenger, arrestin and trafficking readouts each describe one limb of receptor behaviour.
An axis-level readout is not interchangeable with a receptor-level readout from the same study.
Adding receptors to a profile changes the question being asked rather than answering the old one.

Literature density, and why I stopped counting papers

Someone still carrying the original wording, "What is the strongest peptide for fat loss?", will find that density counting looks like the natural substitute for a ranking, and for a while I thought so too. Then I tried it properly. At one point I ranked these compounds by how many papers mention them, and I abandoned the attempt within a week. Raw counts reward age, reward a compound that has been renamed several times, and reward review articles that cite each other. They also reward the kind of paper that repeats a name in an introduction without measuring anything, which is exactly the material I am trying to keep out of my notes.

What I count instead is narrower: how many independent groups have published a measurement on the compound, whether any group outside the original one has reproduced a binding or signalling result, whether an analytical dataset accompanies the sequence, and whether the conditions are reported well enough to reproduce. Those four questions sort the literature far better than a total, and they have the advantage of being answerable from the paper itself.

This is also where the code-named material separates from the named material. For tirzepatide, retatrutide or semaglutide I can find registered study records, replication and public analytical discussion. For ps r3 I can find none of the three, and that absence is the single most important fact in this notebook. It is why the comparison table above carries a blank rather than a number, and why I would rather leave the blank visible than fill it from a neighbour.

What I actually compare when two compounds are placed side by side
What I look for	What it tells me	What it does not tell me
Receptor profile as stated in a methods section	Which receptors were expressed in that preparation	Whether the stated profile is complete or merely unreported
Assay format and readout	Which signalling limb was measured	Anything about the other limbs, or about the preparation as a whole
Cell background and receptor expression level	Whether two numbers can be compared at all	Whether the comparison survives a different background
Study phase as registered in a public registry	How far the compound has been carried in the public record	Whether the reported finding reproduced in an independent group
Number of independent groups publishing a measurement	Whether a finding has left one laboratory	Whether the finding is correct; replication is not proof

Where ps r3, ps t2 and ps cg sit in this map

I put the three codes in the same table as the named compounds deliberately, because that is where readers encounter them and pretending otherwise would be its own kind of dishonesty. In my own filing they sit in a separate column, and the reason is not caution for its own sake. ps r3, ps t2 and ps cg are codes without a deposited sequence I can open, without a receptor profile I can verify, and without an independent replication of any measurement.

What they do have is a naming environment. The codes sit next to named incretin compounds in the same catalogues, and the claims that circulate about them are borrowed from those neighbours, usually without the borrowing being stated anywhere. I handle the specific collision with retatrutide in ps r3 peptide retatrutide, and the sibling codes in ps t2 peptide and ps cg peptide.

So when the question comes back to me as "What is the strongest peptide for fat loss?" my answer is that the named compounds can be compared on receptor profile, study stage and literature density, and the codes cannot yet be compared on any of the three. That is not a refusal; it is the finding. The structural vocabulary I use to keep the two columns apart starts in the structural unit of peptides and proteins, and the pillar that holds it all together is what is ps r3 peptide.

Sources & further reading

Search links into public bibliographic databases; the notebook quotes no paywalled full text.