Which Peptides Gel, and How to Mix Each One

Published September 2026 · Patriot Labs Research Team · ~15 min read

The short answer

Gelling is aggregation that has run to completion. Peptide chains that should stay dissolved as separate molecules instead find one another, stack into β-sheets, and build a network that traps the solvent between them. Three inputs decide whether that happens in a given vial: the sequence, which sets how sticky the molecule is; the concentration, which sets how often chains meet; and the handling, which supplies or withholds the trigger.

Of those three, the only one that changes between one researcher's vial and another's is usually concentration. The sequence is fixed by what was ordered. The handling is a short list of habits covered in the reconstitution guide. Concentration is set entirely by how much bacteriostatic water goes in, which makes diluent volume the single most useful lever available — and the reason a 5 mg vial of one compound needs more water than a 10 mg vial of another.

What a gel actually is

A dissolved peptide stays dissolved because two forces outweigh the attraction between chains: water solvates the polar and charged groups along the backbone and side chains, and like charges on neighbouring molecules repel one another. Remove either force and the chains begin to associate.

Association is not a single event but a progression, and the visible symptoms map onto its stages. Early association produces oligomers too small to see. As they grow, they reach the size where they scatter visible light, and the solution reads as hazy or faintly cloudy. Larger assemblies become visible particulate or strands. At the far end, if the chains are the kind that stack into extended β-sheet ribbons, those ribbons entangle into a continuous space-filling network across the whole volume, and the solution stops flowing. That is a gel. It is the same process that materials scientists deliberately exploit to make peptide hydrogels, running uninvited in a research vial.

The important consequence is that a gel is not a dissolution problem to be solved with more agitation. The molecule has changed its physical state, and in a β-sheet network the change is largely irreversible. Whatever is in that vial is no longer fully the monomeric molecule the certificate of analysis described.

The three inputs that decide it

Sequence. Hydrophobic and aromatic residues are the sticky ones. Aromatic side chains stack against each other, and hydrophobic residues are pushed together by water that would rather bond with itself. A short sequence with a high density of those residues has fewer polar groups to keep it solvated, which is why a 10-residue peptide can be far more aggregation-prone than a 30-residue one. A separate category of stickiness is deliberate: analogues carrying an attached fatty-acid chain are designed to self-associate, because that is how they achieve a long half-life.

Concentration. Aggregation is a collision-driven process, so its rate rises steeply with how many chains occupy a given volume. This is why the same compound can be perfectly clear at 2 mg/mL and set solid at 20 mg/mL, and why the high-mass vials in any catalogue — the bulk sizes and the multi-compound blends — carry more risk than the small ones even when the sequence is identical.

Handling. Three triggers do most of the damage. Mechanical shear, from shaking rather than swirling, unfolds chains at the air–water interface where they aggregate readily. A solution pH near the peptide's isoelectric point strips its net charge and removes the electrostatic repulsion holding chains apart. And heat accelerates everything. The handling guide treats all three in detail; this page assumes they are already being managed and focuses on what is left, which is the compound and the volume.

Gel risk across the catalogue

The table below covers every peptide product in the catalogue. The risk column reflects the compound's tendency to aggregate or gel on reconstitution, not its stability in storage and not its quality — a high-risk entry is a compound that needs a little more care and a little more water, not a compound that is worse. Read the tier assignments as informed expectations from sequence composition and compound class, qualified in the honest limits section below.

Suggested diluent volumes are starting points that keep each compound in a concentration range where clear solutions are typical. They are not the only workable volumes, and they trade against the in-use life of the vial — the same mass spread across more water lasts through fewer withdrawals. The reconstitution calculator will work the arithmetic for any volume.

Built to self-assemble

Retatrutide and cagrilintide are the two compounds in this catalogue most likely to produce a hazy or gelled vial, and the reason is structural rather than accidental. Both carry an attached fatty-acid chain. That lipid is the feature that gives them their long duration, because it binds serum albumin and slows clearance — but a lipid chain on a water-soluble peptide is, by definition, an amphiphile. Amphiphiles associate in water. Work on lipidated GLP-1 analogues has shown that attaching the fatty acid drives oligomerisation and changes aggregation behaviour relative to the unmodified peptide, and that this class assembles into oligomeric micelles and short fibrils in aqueous solution.^1,2

Cagrilintide adds a second reason. It is an analogue of amylin, and native human amylin is one of the most studied amyloid-forming peptides in biology, with a central region around residues 20–29 that will independently form fibrils.³ The medicinal chemistry that produced cagrilintide was explicitly a response to that problem — its developers describe amylin's "high propensity toward the formation of amyloid fibrils" as the central challenge, and the resulting molecule was engineered to be stable.⁴ That engineering works, and cagrilintide is not native amylin. But the family it comes from is the reason it belongs in this group rather than the robust one.

Practically: these are the vials to reconstitute at the generous end, to add diluent to slowly down the wall, and to leave alone afterwards rather than inverting to "help" them along. A faint opalescence in a lipidated analogue immediately after mixing that clears on standing is different from a persistent haze; the latter is the signal to stop.

Aromatic-dense short sequences

Kisspeptin-10 is the clearest documented case in the catalogue. It is only ten residues, but those ten include an unusually high proportion of aromatic and hydrophobic side chains, which leaves relatively little polar surface to keep the molecule solvated. Its β-sheet aggregation has been studied directly: Nielsen and colleagues showed it forms β-sheet-rich amyloid aggregates at neutral and slightly acidic pH, following a lag phase characteristic of nucleated assembly.⁵ The same work found aggregation was completely inhibited by sub-micellar concentrations of certain surfactants — a useful confirmation that the process is interfacial and amphiphile-sensitive, though not a formulation recommendation for a research vial.

AOD-9604 lands in the same group on composition. It is the C-terminal 176–191 fragment of human growth hormone, and fragments cut out of a folded protein frequently expose hydrophobic faces that were buried in the parent structure and have no partner in the free peptide. The practical consequence is a peptide with modest aqueous solubility for its small mass.

Both are 5 mg vials, which helps: even at a generous 2–3 mL the resulting concentration stays low. The failure mode for these two is not usually a full gel but a haze that never quite clears, which is aggregation caught in the middle of the same progression.

The GH-axis analogues

Tesamorelin and the CJC-1295 variants are growth-hormone-releasing factor analogues, a family with a long-standing reputation for occasional cloudiness on reconstitution. GHRH analogues are helical peptides of around 29–44 residues with amphipathic character — a helix with a hydrophobic face is a structure that associates readily with other copies of itself, which is the same geometry that underlies coiled-coil assembly.

In practice these behave well most of the time and this is a moderate rather than elevated tier. Where they cause trouble it is usually one of two things: cold diluent hitting cold powder, or a vial that was inverted vigorously instead of swirled. Letting both the vial and the bacteriostatic water reach room temperature before mixing removes most of the problem, and is worth the ten minutes it costs.

The blends and the concentration ceiling

The KLOW and Glow blends deserve their elevated tier for a reason that has nothing to do with their ingredients being fragile. Individually, KPV, GHK-Cu, TB-500 and BPC-157 are all in the low-risk tier. The issue is arithmetic: KLOW puts 80 mg of total peptide into a single vial and Glow puts 70 mg. Reconstituted into 1 mL, that is 80 and 70 mg/mL respectively — concentrations at which the collision rate between chains is high enough that even well-behaved sequences start associating, and at which any one component approaching its own solubility limit will come out of solution and seed the rest.

A second consideration applies specifically to these two: both contain GHK-Cu, and a copper-carrying peptide in a vial with other peptides is a chemistry question in its own right, treated at length in the pH and mixing guide. That guide's conclusion is worth repeating here — a manufactured blend, formulated and lyophilised together, is not the same thing as combining two vials on a bench, and the former has been through a process the latter has not.

The fix for the blends is simply more water. At 4–5 mL, KLOW lands between 16 and 20 mg/mL, which is an ordinary working concentration rather than an extreme one.

The robust majority

Most of the catalogue does not gel, and it is worth being clear about why, because the reasons are as instructive as the failures.

Very short peptides cannot build a network. KPV and Pinealon are tripeptides; Epithalon is a tetrapeptide; SS-31 is four residues. A β-sheet network needs chains long enough to form extended hydrogen-bonded ribbons that entangle. Three or four residues cannot do it, which is why Epithalon tolerates 50 mg in a vial without difficulty while a 50 mg vial of a long amphipathic peptide would be a different proposition.

Strong net charge keeps chains apart. SS-31, FOXO4-DRI, Semax and Selank all carry substantial charge at ordinary solution pH. Electrostatic repulsion is the force that keeps dissolved molecules dissolved, and a peptide sitting far from its isoelectric point has plenty of it.

Cyclic structures resist stacking. PT-141 and Melanotan II are both constrained rings. β-sheet formation requires a chain to adopt an extended conformation, and a lactam bridge simply will not allow it.

Some sequences are just soluble. BPC-157 and TB-500 are the two most reliably clear compounds in the catalogue, which is fortunate given how often they appear in blends. GHK-Cu and AHK-Cu carry 100 mg per vial and still dissolve readily, because a charged copper-binding tripeptide is about as water-friendly as peptides get. For those two, the handling concern is the copper centre and its interactions, not gelation.

The ones that are not peptides at all

Four catalogue items are routinely reconstituted alongside peptides but follow different chemistry entirely, and applying peptide logic to them gives the wrong answer.

NAD+ is a dinucleotide coenzyme, not a peptide. It will not form β-sheets and cannot gel in the sense this guide describes. Its failure modes are hydrolysis and pH sensitivity; a 1000 mg vial is about solubility and stability, not aggregation. 5-Amino-1MQ and SLU-PP-332 are small molecules, where the relevant limit is simple solubility — a small molecule either dissolves at a given concentration or it does not, and the result is precipitate rather than gel. Glutathione is technically a tripeptide but behaves as a small molecule; its real enemy is oxidation, since its free thiol is what makes it useful and what makes it fragile.

The proteins — HCG, HMG and botulinum toxin complex — are a separate category again. These are large folded structures, and their failure mode is denaturation at air–water interfaces followed by aggregation of the unfolded protein. That is why the universal instruction for all three is to add diluent slowly down the wall and never, under any circumstances, to shake. A protein solution that has been shaken into foam has lost material to that foam, and no amount of standing will bring it back.

Reconstitution volumes, worked out

Because concentration is the lever, it helps to see what the common volumes actually produce. The table below is straightforward division — vial mass divided by diluent volume — for the compounds where the number matters most.

The pattern worth internalising is that doubling the diluent halves the concentration, and that the difference between a 1 mL and a 3 mL reconstitution is a threefold change in how often chains meet. For a compound with no gelling tendency that choice is purely about convenience. For retatrutide at 60 mg it is the difference between 60 mg/mL and 20 mg/mL, which is a different physical regime.

What to do with a vial that has already gelled

The honest answer is that a gelled vial should be treated as compromised and not used for research work. β-sheet aggregation is largely irreversible, and the visible gel is the endpoint of a process that has been running since the diluent went in. Warming the vial or agitating it may make the contents look more liquid, but restoring the appearance is not the same as restoring the molecule. Some fraction of the material is now in an aggregated state, that fraction is unknown, and the solution no longer matches the certificate of analysis in either concentration or composition. Data generated from it is data of unknown provenance.

What is worth doing is diagnosing it, so the next vial behaves. Ask, in order: how much diluent went in, and was it at the low end for that compound? Was the water added down the wall or squirted onto the cake? Was the vial swirled or shaken? Were both the powder and the diluent at room temperature? Is this a compound in the elevated tier above? In the large majority of cases one of those five questions produces the answer, and the correction is straightforward.

If a compound gels repeatedly at a volume that should be comfortable for it, that is worth reporting rather than working around — unexpected physical behaviour in a sequence that should be robust is exactly the kind of signal that belongs with the batch's certificate of analysis.

What this guide does not establish

The risk tiers on this page are reasoned from sequence composition, compound class, and the published behaviour of closely related molecules. They are not the output of per-product stability studies on Patriot Labs material, and they should not be read as one. Three entries rest on direct literature for the specific molecule or its immediate parent — kisspeptin-10, the amylin family behind cagrilintide, and the lipidated GLP-1 class that retatrutide belongs to. The remainder are informed expectations.

Suggested diluent volumes are conservative starting points, not specifications, and they carry no claim about what concentration is appropriate for any particular experiment. Individual batches vary, and a compound in the low-risk tier can still produce a hazy vial if it is mixed cold, shaken, or reconstituted at an unusual concentration. Sequence-level properties such as isoelectric point are discussed qualitatively here; the pH guide treats them quantitatively.

Finally, nothing on this page describes or implies any use of these materials outside a laboratory setting.

Frequently asked questions

Why did my peptide turn into a gel?

A gel is aggregation that has gone far enough to build a space-filling network. Instead of staying separated in solution, peptide chains associate into β-sheet structures that trap solvent between them. Three things drive it together: a sequence with enough hydrophobic or aromatic residues to stick, a concentration high enough for chains to find each other, and a trigger such as mechanical shear, a pH near the peptide's isoelectric point, or heat. The most common avoidable cause is simply too little diluent.

Which research peptides are most likely to gel?

Three groups account for most cases. Lipidated analogues carrying a fatty-acid chain — retatrutide and cagrilintide here — self-assemble by design. Short sequences dense in aromatic and hydrophobic residues, of which kisspeptin-10 is the clearest example, form β-sheet aggregates readily. And high-mass blends reconstituted into a small volume reach concentrations where almost any sequence becomes more prone to association.

Can a gelled peptide vial be recovered?

Generally no, and it should not be treated as recoverable. Warming or agitating may improve the appearance, but β-sheet aggregation is largely irreversible and redissolving the visible gel does not restore the original monomeric molecule. The material is no longer fully what the certificate of analysis described, so its composition is unknown. Discard it and reconstitute a fresh vial at a lower concentration.

Does adding more bacteriostatic water prevent gelling?

Lower concentration reduces how often peptide chains encounter one another, so for a sequence with known gelling behaviour a larger diluent volume is one of the few levers available, and often the difference between a clear solution and a hazy one. It is not a guarantee — sequence, pH and handling still matter — and it trades against the in-use life of the vial, since the same mass is spread across more volume.

Is a cloudy peptide the same as a gelled peptide?

They are the same process at different severities. Mild aggregation produces particles large enough to scatter light, which reads as haze. Heavier aggregation produces visible particulate, and in the extreme the chains build a continuous network and the solution sets. Cloudiness is the early warning that the same mechanism is underway.

Does a gel mean the peptide was low quality?

Not by itself. Gelling is a physical behaviour of the sequence under the conditions in the vial, and several high-purity compounds aggregate readily precisely because of what they are — a lipidated analogue is designed to self-associate. Purity and aggregation propensity are different properties. Repeated gelling of a compound that should be robust is a different matter and worth raising.

References

1. Effect of Lipidation on the Structure, Oligomerization, and Aggregation of Glucagon-like Peptide 1. Bioconjugate Chemistry (2025), 36(3):401. PMID 39841169; doi:10.1021/acs.bioconjchem.4c00484. Establishes that attaching a fatty-acid chain alters oligomerisation and aggregation behaviour relative to the unmodified peptide.
2. Semaglutide Aggregates into Oligomeric Micelles and Short Fibrils in Aqueous Solution. PubMed Central PMC12152837. Direct observation of micellar and fibrillar assembly in a lipidated GLP-1 analogue in aqueous solution.
3. Alghrably, M., Bennici, G., Szczupaj, G., et al. Exploring the central region of amylin and its analogs aggregation: the influence of metal ions and residue substitutions. Frontiers in Chemistry (2024), 12:1419019. doi:10.3389/fchem.2024.1419019. Identifies residues 20–29 as the amyloidogenic core of amylin and shows how substitutions and metal ions modulate fibril formation.
4. Kruse, T., Hansen, J.L., Dahl, K., et al. Development of Cagrilintide, a Long-Acting Amylin Analogue. Journal of Medicinal Chemistry (2021), 64(15):11183–11194. doi:10.1021/acs.jmedchem.1c00565. Describes amylin's high propensity toward amyloid fibril formation as the central design challenge and cagrilintide as the stable, lipidated result.
5. Nielsen, S.B., Franzmann, M., Basaiawmoit, R.V., Wimmer, R., Mikkelsen, J.D., & Otzen, D.E. Beta-sheet aggregation of kisspeptin-10 is stimulated by heparin but inhibited by amphiphiles. Biopolymers (2010), 93(8):678–689. doi:10.1002/bip.21434; PMID 20301214. Direct study of β-sheet amyloid aggregation in kisspeptin-10, including its lag phase and inhibition by sub-micellar surfactants.
6. Mechanical characteristics of β-sheet-forming peptide hydrogels are dependent on peptide sequence, concentration and buffer composition. Royal Society Open Science (2018), 5:171562. Demonstrates that concentration and buffer conditions, not sequence alone, control gelation.

For the underlying chemistry in more depth, see the companion guides on peptide handling best practices, how peptides degrade, pH and what happens when two peptides share a vial, and how to reconstitute research peptides.

All Patriot Labs products are sold strictly for in-vitro research and laboratory use only. Not for human or veterinary consumption. This guide is educational and describes peptide aggregation chemistry in general terms; it is not medical advice, does not describe how to use any product, and the reconstitution volumes cited are general starting points rather than a specification for any individual compound.