# Peptide testing — deep dive

**Phase:** Deep-dive (round 2)
**Last updated:** 2026-04-23
**Sits alongside:** `research/summary.md`, `research/01-existing-services/finnrick.md`, `research/02-analytical-chemistry/hplc.md`, `research/02-analytical-chemistry/lc-ms.md`, `research/06-reference-standards/peptide-standards.md`

---

## Key takeaways

- **Peptides are not small molecules.** They are 500–6,000 Da, they ionize differently, most have weak or no UV chromophores at 280 nm, and HPLC-UV alone **cannot** confirm sequence. A truncated or scrambled analog can share retention time and UV spectrum with the real thing.
- **BPC-157 has zero aromatic residues (no Trp/Tyr/Phe).** 280 nm detection simply does not work. Anyone quoting a "280 nm purity %" for BPC-157 is either running at 214 nm and mislabeling or measuring noise. Same failure mode for TB-500 (heptapeptide fragment) and several nootropic peptides.
- **Sequence confirmation requires LC-MS/MS**, not single-quad. Intact mass from a benchtop single-quad (~$25–35k used) will catch wrong-molecule substitution but not an N-terminal truncation or a residue swap that shifts mass by <2 Da. Community labs doing real sequence work — MZ Biolabs, Janoshik, Krause — run ion traps or Q-TOFs.
- **Peptide testing costs 2–5× more than steroid testing.** Janoshik charges $300 for a GLP-1 peptide vs $120 for an anabolic oil. Reference standards are $500–2,000/mg where they exist at all; for newer peptides like retatrutide, no USP RS exists and working standards come from Bachem or custom synthesis houses.
- **Finnrick's dataset shows ~1-in-3 gray-market peptides fail basic expectations; ~8% have quantifiable endotoxin above their internal 40 EU/vial threshold.** GLP-1s are the most-tested class (tirzepatide 1,776 samples, retatrutide 2,489, semaglutide 309 as of 2026-04-23).

---

## 1. Why peptides are different

The round-1 HRT docs focus on molecules like estradiol valerate (MW 356.5, C₂₃H₃₂O₃) and testosterone cypionate (MW 412.6, C₂₇H₄₀O₃). Everything downstream — column choice, detector wavelength, ionization mode, reference-standard sourcing, reference-library availability — is calibrated around small, neutral, UV-active steroids. None of that transfers directly to peptides.

### 1.1 Size and physicochemistry

| Class | Typical MW | Aromatic chromophore? | Ionization |
|---|---|---|---|
| Steroid HRT (EV, EC, T-cyp) | 300–420 Da | Phenolic A-ring (E2 esters) or neutral conjugated dienone (T esters) | ESI neutral, often poor |
| BPC-157 | 1,419.55 Da (avg) | **None** | ESI+, multiply charged |
| Ipamorelin | 711.85 Da | D-2-Nal + D-Phe | ESI+ |
| Melanotan-II | 1,024.18 Da | Trp + D-Phe (cyclic) | ESI+ |
| CJC-1295 (no DAC) | 3,367.80 Da | 2× Tyr + Phe | ESI+, [M+3H]³⁺ |
| CJC-1295 + DAC | 3,647.31 Da | 2× Tyr + Phe + maleimide | ESI+, [M+3H]³⁺ |
| Semaglutide | 4,113.58 Da | Trp, Tyr, Phe | ESI+, [M+4H]⁴⁺ common |
| Tirzepatide | 4,813.45 Da | Trp, 2× Tyr, Phe | ESI+, [M+4H]⁴⁺ / [M+5H]⁵⁺ |
| Retatrutide | 4,731.33 Da | 3× Tyr, 2× Phe (no Trp) | ESI+ |
| Tesamorelin | 5,135.86 Da | 2× Tyr + Phe (no Trp) | ESI+ |

**Claim:** BPC-157 is a 15-residue pentadecapeptide with sequence Gly-Glu-Pro-Pro-Pro-Gly-Lys-Pro-Ala-Asp-Asp-Ala-Gly-Leu-Val and average molecular weight 1419.55 Da (C₆₂H₉₈N₁₆O₂₂).
- **Confidence:** C1
- **Source:** https://pubchem.ncbi.nlm.nih.gov/compound/Bpc-157 ; https://pubmed.ncbi.nlm.nih.gov/21548867/
- **Date checked:** 2026-04-23
- **Notes:** Sequence has **no aromatic side chains**. Trp/Tyr/Phe are absent. Any 280 nm method for BPC-157 is invalid.

**Claim:** Semaglutide (C₁₈₇H₂₉₁N₄₅O₅₉, 4113.58 Da avg) is a 31-residue GLP-1 analog with Trp at position 31 (C-terminal), Tyr at position 13, and Phe at position 6; the lysine at position 26 carries a Nε-γGlu-(AEEA)₂-C18-diacid modification.
- **Confidence:** C1
- **Source:** https://pubchem.ncbi.nlm.nih.gov/compound/Semaglutide ; https://en.wikipedia.org/wiki/Semaglutide
- **Date checked:** 2026-04-23

**Claim:** Tirzepatide (C₂₂₅H₃₄₈N₄₈O₆₈, 4813.45 Da) is a 39-residue dual GIP/GLP-1 agonist with sequence H-Tyr-Aib-Glu-Gly-Thr-Phe-Thr-Ser-Asp-Tyr-Ser-Ile-Aib-Leu-Asp-Lys-Ile-Ala-Gln-Lys(γGlu-C20 diacid)-Ala-Phe-Val-Gln-Trp-Leu-Ile-Ala-Gly-Gly-Pro-Ser-Ser-Gly-Ala-Pro-Pro-Pro-Ser-NH₂.
- **Confidence:** C1
- **Source:** https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?tab=structure&ligandId=11429 ; https://en.wikipedia.org/wiki/Tirzepatide
- **Date checked:** 2026-04-23
- **Notes:** Despite common loose statements, tirzepatide **does contain Trp** (position 25) plus 2× Tyr and 2× Phe. 280 nm detection works well.

**Claim:** Retatrutide (C₂₂₁H₃₄₂N₄₆O₆₈, 4731.33 Da) is a 39-residue triple GIP/GLP-1/glucagon agonist (Lilly LY3437943). No Trp; has 3× Tyr and 2× Phe.
- **Confidence:** C1
- **Source:** https://pubchem.ncbi.nlm.nih.gov/compound/Retatrutide ; https://gsrs.ncats.nih.gov/ginas/app/beta/substances/NOP2Y096GV
- **Date checked:** 2026-04-23
- **Notes:** Common community-forum formula "C₂₃₃..." is wrong; correct is C₂₂₁. All current gray-market material is pre-approval; no USP RS exists.

### 1.2 Stability and storage

Peptides fail in ways that steroid esters do not:

- **Oxidation** — Met and Trp are the two most-oxidized residues. +16 Da (Met-sulfoxide, Trp-oxindole) +32 Da (sulfone, N-formylkynurenine). Semaglutide's Trp³¹ is a known degradation hotspot. Tirzepatide's Trp²⁵ is similar. A vial that oxidized in transit reads as lower purity by HPLC.
- **Deamidation** — Asn and Gln → Asp and Glu (+1 Da). Under aqueous reconstitution at pH > 6, half-life can be days to weeks. BPC-157 has no Asn but two Asp already; not a major degradation path. Semaglutide has multiple Gln/Asn residues.
- **Disulfide scramble** — peptides with Cys can form intra- or intermolecular disulfides in storage. None of the popular GLP-1 analogs have Cys, but some legacy peptides (oxytocin, somatostatin, desmopressin) do.
- **Aggregation** — dimers and higher n-mers appear as later-eluting peaks (hydrophobic) or as off-specification "late eluting polar impurities." Semaglutide fibrillates at concentrations >10 mg/mL under stress.
- **Adsorption** — peptides adsorb to plastic and glass surfaces. A 0.5 mg/mL tirzepatide solution in a polypropylene vial can lose >10% to wall adsorption within 24 h without a surfactant (polysorbate 20 or 80 is the pharmaceutical fix). This matters for home-lab dilution series.

### 1.3 Sequence confirmation — the key epistemic gap

HPLC-UV can tell you: "something of approximately this hydrophobicity elutes at approximately this retention time and absorbs at approximately this wavelength." That's it.

It **cannot** tell you:
- Whether the N-terminus is intact (truncations often elute within 30 s of parent)
- Whether D/L stereochemistry is correct
- Whether a single-residue substitution has been made (e.g., Leu → Ile — identical mass; Ala → Gly — −14 Da)
- Whether the correct peptide is present if an isobaric scrambled analog coelutes

For those, you need **tandem mass spectrometry** — MS/MS with CID or HCD fragmentation to read b/y ion ladders. See §3.

### 1.4 Impurity profile (what HPLC actually sees)

Typical peptide-synthesis impurities in a purified SPPS product:
- **Truncation products** (−1, −2, −3 residues): elute earlier or later depending on the removed residue; mass differs by the residue weight (71–186 Da).
- **Deletions** (internal residue missing): similar mass shift, different retention.
- **Oxidations** (+16, +32 Da, Met/Trp/Cys).
- **Deamidations** (+0.98 Da) — visible only in high-res MS, not in unit-resolution MS.
- **Racemizations** (same mass, different retention).
- **Aspartimide formation** (Asn, Asp in sequence): −18 Da then reopens to α- or β-Asp.
- **Incomplete side-chain deprotection** (tBu +56 Da, Trt +242 Da).
- **Free fatty-acid side-chain conjugate hydrolysis** — for semaglutide, tirzepatide, retatrutide: loss of the C18/C20 diacid chain. Mass drop 300–400 Da, dramatic retention change.
- **Aggregates / dimers** (2×, 3× MW).
- **Salts / counter-ions** (TFA, acetate): affect net peptide content (NPC) — a vial listed as "10 mg" of TFA-salt semaglutide is only ~8 mg of free-base peptide.

---

## 2. UV detection strategy

Peptide absorbance is a function of backbone amides (peptide bond) and aromatic side chains.

| Wavelength | What absorbs | ε per residue (M⁻¹ cm⁻¹) | Pros | Cons |
|---|---|---|---|---|
| **214 nm** | Peptide bond (n→π*) | ~1,000–10,000 per bond | Universal — detects every peptide regardless of aromatics | TFA (very strong 214 absorber), ACN cutoff ~190 nm ok; baseline drift on gradient; lower dynamic range |
| 220 nm | Peptide bond shoulder | ~600–3,000 per bond | Slightly cleaner baseline than 214 | Sensitivity ~3–5× lower |
| **280 nm** | Trp (ε ≈ 5,500), Tyr (ε ≈ 1,400), Phe (ε ≈ 200) | Depends on content | Clean baseline; solvent-agnostic | Useless for peptides without aromatics (BPC-157, TB-500 short, some GHKs) |
| 254 nm | Incidental; select modified residues | varies | Cheap fixed-wavelength lamps | Rarely applicable |

### 2.1 Mobile phase trap: TFA vs FA

The default peptide ion-pair reagent is **0.1% trifluoroacetic acid (TFA)** in water/ACN. It gives the sharpest peak shape on reversed-phase C18/C4/C8 columns — crucial for resolving closely-eluting impurities. But:

- TFA absorbs strongly at 210–220 nm. A well-equilibrated gradient still produces noticeable baseline rise when %B climbs.
- TFA ion-suppresses ESI+. Typical loss 5–20× in MS signal vs formic acid for routine conditions; the extreme end (up to ~50×) is only reached with supercharging or very high TFA loadings.

**Practical compromise:**
- HPLC-UV-only work: 0.1% TFA, detect at 214 nm. Accept the baseline drift.
- LC-MS work: 0.1% formic acid (FA), or 0.05% TFA + 0.05% FA ("low-TFA"), detect at 214 or 220 nm.
- Avoid: phosphate buffers (non-volatile, trash your MS source) or ammonium acetate at high-pH mobile phase (Tyr/Trp pKa shifts move retention).

**Implication for a home lab:** a VWD (variable wavelength detector) set at 214 nm will run every peptide on your list, but your LOD/LOQ will be 3–10× worse than a DAD system that does 214 + 280 dual-channel acquisition. Method development for peptide gradients at 214 nm is less forgiving than at 280 nm — every TFA batch difference, every tiny pH drift, shows up as baseline noise.

### 2.2 Quantification at 214 nm

A commonly-cited rule-of-thumb (Kuipers & Gruppen 2007; corrected in round 2, R2-M9, from an earlier multiplicative form): molar absorptivity at 214 nm is **additive**, not multiplicative:

```
ε₂₁₄  ≈  923 × (n − 1)  +  8,400 · N_Trp  +  1,400 · N_Tyr  +  (smaller terms for Phe, His, Cys, Met, disulfides)
```

where n is the residue count and the first term is the backbone peptide-bond contribution. In practice peptide quantification at 214 nm by **area normalization** (purity %) assumes all impurities have the same ε₂₁₄ as the parent — which is approximately true if all impurities are peptide-bond-based (truncations, deletions). It is **false** when:
- The impurity is a small molecule (TFA counter-ion, residual scavenger like TIS).
- The impurity has a different chromophore (oxidized Trp → N-formylkynurenine, ε shifts).
- The impurity is a fatty-acid conjugate fragment with no peptide bonds.

This is why Finnrick's methodology page and all partner-lab questionnaires emphasize **quantification against a reference standard** rather than purity-normalized signal.

---

## 3. Sequence confirmation via LC-MS/MS

### 3.1 Basics

A peptide ionized by ESI+ typically gives **multiply charged ions**: [M+2H]²⁺, [M+3H]³⁺, [M+4H]⁴⁺ for larger sequences. Charge-state envelope analysis is the first-pass identity check — you need the right set of m/z values at the right spacing.

For a 4,113 Da semaglutide:
- [M+2H]²⁺ = 2057.79 m/z
- [M+3H]³⁺ = 1372.20 m/z
- [M+4H]⁴⁺ = 1029.40 m/z
- [M+5H]⁵⁺ = 823.72 m/z

Deconvolution gives 4,113.58. Great — you've confirmed a 4,113 Da molecule eluted. But **4,113 Da matches roughly 50 known semaglutide-adjacent sequences** including racemates, Aib-Ala swaps, des-amide analogs. Mass alone does not equal identity.

### 3.2 Tandem MS: reading the sequence

Collision-induced dissociation (CID) in an ion trap or collision cell breaks amide bonds preferentially. Two main fragment series:
- **b-ions** — N-terminal fragments + C=O (retain N-terminus, lose C-terminus residues).
- **y-ions** — C-terminal fragments + H (retain C-terminus, lose N-terminus residues).

A b/y ladder reads the sequence one residue at a time, N-to-C or C-to-N. For a clean 31-mer like semaglutide, CID on the [M+4H]⁴⁺ gives a usable ladder covering most of the sequence. The Trp at position 31 and Tyr at 13 are useful "anchor" residues that are easy to identify by their diagnostic immonium ions (Trp 159 m/z, Tyr 136 m/z).

### 3.3 Instrument tiers for MS/MS

| Tier | Example models | Used price band | Can de novo sequence? | Notes |
|---|---|---|---|---|
| Single quadrupole | Agilent 6120, Shimadzu LCMS-2020, Thermo ISQ EC | $25–50k | **No** | Intact mass only. Will catch wrong-molecule substitution (e.g., liraglutide sold as semaglutide). Will **not** catch a sequence variant of equal mass. |
| Triple quadrupole | Agilent 6460/6470, Sciex 4000/5000, Waters Xevo TQ | $35–120k | No (targeted MRM only) | Quantitative workhorse. Can do targeted MRM on known peptide transitions, but can't read a sequence you don't already know. |
| Linear ion trap | Thermo LTQ (Classic, XL, Velos, Velos Pro) | $40–100k | **Yes** | Classical peptide MS/MS platform. MZ Biolabs runs an LTQ Velos Pro. CID fragmentation with MSⁿ capability. |
| Q-TOF | Bruker Compact, Waters Synapt / Xevo G2, Agilent 6530/6545 | $60–200k | **Yes**, with high mass accuracy (<5 ppm) | MZ Biolabs runs 3× Bruker Compact QTOF. Gives formula-level confirmation; essential for modified peptides. |
| Orbitrap | Thermo Q Exactive, Fusion, Eclipse | $100k+ (often >$250k) | Yes, with ~1 ppm | Top-of-line for peptide PTM work; rarely economic for a community lab. |

**Claim:** MZ Biolabs operates 3× Bruker Compact QTOF mass spectrometers paired with Waters Acquity UPLC, plus a Thermo LTQ Velos Pro paired with a Waters nanoAcquity. Peptide sequence confirmation uses optional MS² fragmentation on these platforms.
- **Confidence:** C1
- **Source:** https://www.mzbiolabs.com/mzbiolabs/our-techniques/
- **Date checked:** 2026-04-23
- **Notes:** nanoAcquity flow rate is 100 nL/min to 5 µL/min, targeting 5 pg/mL detection for PK work. Not all Finnrick samples at MZ get MS²; it is an optional add-on per the page text.

### 3.4 Home-lab MS infeasibility for sequence work

A used single-quad LC-MS at $25–35k will confirm **intact mass** — useful but necessary-not-sufficient. To read sequence you need an ion trap or Q-TOF. Real-world used pricing:

- **Thermo LTQ classic / XL**: $25–60k used (LabX / BioSurplus). Parts availability is getting worse; Thermo stopped PM contracts on Classic/XL years ago. Velos and Velos Pro run $50–100k used and still have aftermarket support.
- **Bruker amaZon / HCT Ultra (ion trap)**: $30–70k used; similar aftermarket caveat.
- **Agilent 6520 / 6530 Q-TOF**: $60–150k used.
- **Waters Synapt G2 / G2-S**: $80–180k used.

Add $10–20k for a nitrogen generator, vacuum roughing pumps on service, and CDS software keys that often **do not transfer** on re-sale. Total realistic home installation: $60–120k for a used ion trap LC-MS/MS plus LC stack — plus $5–15k/year in consumables (N₂, service parts, turbo rebuilds).

The honest recommendation: **outsource MS/MS to MZ Biolabs or Janoshik.** Run HPLC-UV at home for purity screening; pay for orthogonal MS confirmation per-sample or per-batch.

---

## 4. Per-peptide detail table

For each of the peptides that appear meaningfully in Finnrick / Janoshik / MZ Biolabs workflows:

### 4.1 Semaglutide

- **Formula / MW:** C₁₈₇H₂₉₁N₄₅O₅₉ / 4113.58 Da avg, 4111.13 mono
- **Sequence (31-mer, Aib⁸, Arg³⁴, Lys²⁶(γGlu-AEEA-AEEA-C18diacid)):** HAibEGTFTSDVSSYLEGQAAK(mod)EFIAWLVRGRG
- **Aromatics:** Phe⁶, Tyr¹³, Trp³¹ → strong 280 nm
- **HPLC column:** C18 (ACE UltraCore SuperC18 2.6 µm; Waters BEH300 C18 2.5 µm; Agilent AdvanceBio Peptide Mapping) or C8 at higher pH; column temp 40–60 °C for peak sharpness
- **Mobile phase:** A: 0.1% TFA / water; B: 0.1% TFA / ACN. Gradient 20–60% B over 20 min for full profiling
- **MS adducts:** [M+3H]³⁺ = 1372.2, [M+4H]⁴⁺ = 1029.4
- **Key impurities:** Aib⁸-racemate; Lys²⁶-side-chain fatty-acid hydrolysis products (des-diacid); Met-sulfoxide (no Met in sequence, so this is N/A); Trp³¹ oxidation (+16 Da); deamidation products
- **Gray-market failure modes:** under-dosing (common per Finnrick ratings); occasional sequence substitution (liraglutide-for-semaglutide historically reported in counterfeit Ozempic; FDA public alerts)
- **Finnrick test count:** 309 samples / 51 vendors (2026-04-23)

### 4.2 Tirzepatide

- **Formula / MW:** C₂₂₅H₃₄₈N₄₈O₆₈ / 4813.45 Da avg
- **Sequence (39-mer, Aib², Aib¹³, Lys²⁰(γGlu-C20 diacid)):** YAibEGTFTSDYSIAibLDKIAQK(mod)AFVQWLIAGGPSSGAPPPS-NH₂
- **Aromatics:** Tyr¹, Phe⁶, Tyr¹⁰, Phe²², Trp²⁵ → strong 280 nm
- **HPLC column:** Same as semaglutide. C18 or phenyl-hexyl. Lilly reference method uses a C18 at 65 °C
- **MS adducts:** [M+4H]⁴⁺ = 1204.4, [M+5H]⁵⁺ = 963.7
- **Key impurities:** D-Tyr¹ (racemization at N-terminus is common); Aib²→Ala substitution (−14 Da); fatty-diacid des-modification; oxidation of Trp²⁵
- **Gray-market failure modes:** under-dosing dominant; on Finnrick data, tirzepatide under-dose is the #1 reason for quantity-accuracy score failures (absolute number of tested vials is high so failure rate can be large in absolute terms even if percentage is modest)
- **Finnrick test count:** 1,776 samples / 136 vendors

### 4.3 Retatrutide

- **Formula / MW:** C₂₂₁H₃₄₂N₄₆O₆₈ / 4731.33 Da avg
- **Sequence (39-mer, Aib², αMeLeu¹³, Aib²⁰, Lys¹⁷(γGlu-AEEA-C20 diacid)):** YAibQGTFTSDYSI(αMeLeu)LDKK(mod)AQAibAFIEYLLEGGPSSGAPPPS-NH₂
- **Aromatics:** Tyr¹, Phe⁶, Tyr¹⁰, Phe²², Tyr²⁵ — **no Trp** → moderate 280 nm (Tyr dominant, ε ≈ 1400 × 3 = 4200) (R2 minor: aromatic positions off by one in earlier draft)
- **HPLC column:** C18 ~ 6.19 min retention per Finnrick example COA (specific gradient not disclosed but typical 25–55% B over 10 min at 40 °C)
- **MS adducts:** [M+4H]⁴⁺ = 1183.8, [M+5H]⁵⁺ = 947.3
- **Key impurities:** αMeLeu¹³ synthetic impurities (this is a non-natural residue so SPPS purity varies); racemizations at Tyr¹ and Asp⁹; Aib racemization; incomplete fatty-acid conjugation
- **Gray-market failure modes:** **most-tested peptide on Finnrick** (2,489 samples / 164 vendors) — highest absolute failure count, driven by the fact that retatrutide is still in clinical trial at Lilly (not commercially approved) so every gray-market sample is pre-commercial and vendor quality control is uneven. No USP RS; reference peptides come from Bachem, CS Bio, or China CROs.
- **Finnrick test count:** 2,489 samples / 164 vendors

### 4.4 BPC-157 (critical peptide-specific pitfall)

- **Formula / MW:** C₆₂H₉₈N₁₆O₂₂ / 1419.55 Da avg, 1418.72 mono
- **Sequence (15-mer):** Gly-Glu-Pro-Pro-Pro-Gly-Lys-Pro-Ala-Asp-Asp-Ala-Gly-Leu-Val (GEPPPGKPADDAGLV)
- **Aromatics:** **None.** No Trp, no Tyr, no Phe.
- **Critical:** 280 nm detection cannot detect BPC-157. Period. Any COA reporting a 280 nm purity number for BPC-157 is either (a) actually measuring at 214 nm and mislabeling, (b) measuring an aromatic impurity rather than BPC-157, or (c) fraudulent.
- **HPLC column:** C18 at low pH; BPC-157 elutes early because of its hydrophilic core (3× Asp, 2× Glu, 1× Lys). Typical: 5–30% B over 15 min, retention ~6–10 min
- **MS adducts:** [M+H]⁺ = 1419.7 (ion trap readable); [M+2H]²⁺ = 710.4
- **Key impurities:** Asp-to-isoAsp rearrangement at Asp¹⁰ / Asp¹¹ (common for Asp-Asp sequences); aspartimide cyclization; Lys⁷ side-chain modifications from incomplete deprotection
- **Gray-market failure modes:** Wide variation in purity and potency. Finnrick's BPC-157 dataset (501 samples / 72 vendors) includes multiple D and F ratings. Because any lab using only 280 nm detection on BPC-157 is producing invalid numbers, historical COAs from gray-market vendors are particularly unreliable — much of the published "99%" purity on BPC-157 vials came from inadequate methodology before Finnrick-partnered labs made 214 nm standard.
- **Finnrick test count:** 501 samples / 72 vendors

### 4.5 TB-500 / thymosin β4

**Ambiguity note:** "TB-500" has two meanings in the gray market:
- **Short form** (most common in gray-market vials): Ac-Leu-Lys-Lys-Thr-Glu-Thr-Gln (7-mer, ~889 Da), corresponding to positions 17–23 of thymosin β4. Sometimes just called "LKKTETQ." **Zero aromatics.**
- **Full thymosin β4** (43-mer, ~4963 Da) with one Phe at position 12. Some Trp/Tyr absent. Weak 280 nm.

Either way, 214 nm is required. Gray-market "TB-500" is often the short form sold at full-length prices; LC-MS intact-mass is the only way to tell.

- **HPLC column:** C18, shallow gradient; short form elutes around 30% B
- **MS adducts:** Short [M+H]⁺ = 890.5; full Tβ4 [M+3H]³⁺ = 1655.3
- **Finnrick test count:** 83 samples / 11 vendors

### 4.6 CJC-1295 (with and without DAC)

- **Without DAC:** C₁₅₂H₂₅₂N₄₄O₄₂ / 3367.80 Da; 30-mer modified GHRH. Sequence H-Tyr-D-Ala-Asp-Ala-Ile-Phe-Thr-Gln-Ser-Tyr-Arg-Lys-Val-Leu-Ala-Gln-Leu-Ser-Ala-Arg-Lys-Leu-Leu-Gln-Asp-Ile-**Leu**-Ser-Arg-NH₂ (round-2 correction R2-M3: position 27 is **Leu**, not Met — the whole point of the modified-GRF backbone is D-Ala²/Gln⁸/Ala¹⁵/**Leu²⁷** to eliminate the Met-oxidation vulnerability present in native sermorelin/GHRH)
- **With DAC:** 3647.31 Da. Same sequence with Nε-maleimidopropionyl-Lys³⁰ appended (the DAC = Drug Affinity Complex binds serum albumin via Cys³⁴ Michael addition, extending half-life)
- **Aromatics:** Tyr¹, Phe⁶, Tyr¹⁰ → moderate-strong 280 nm
- **HPLC column:** C18, 20–55% B over 20 min
- **MS adducts:** no-DAC [M+3H]³⁺ = 1123.6; with-DAC [M+3H]³⁺ = 1216.8
- **Key impurities:** D-Ala²/L-Ala² racemization; maleimide hydrolysis (DAC variant); truncation -Arg²⁹-NH₂; incomplete N-terminal substitution (R2-M3 — Met²⁷ oxidation removed because CJC-1295 has Leu²⁷ not Met²⁷; that impurity target belongs on sermorelin, not CJC-1295)
- **Gray-market failure modes:** **Mislabeling "CJC-1295 with DAC" vs "without DAC" is the #1 failure mode** — 280 Da mass difference is trivial on MS but invisible on UV. Many vendors mix blends (CJC/Ipamorelin) where the CJC content is actually the cheaper no-DAC variant
- **Finnrick test count:** 248 samples / 47 vendors

### 4.7 Ipamorelin

- **Formula / MW:** C₃₈H₄₉N₉O₅ / 711.85 Da
- **Sequence (pentapeptide):** Aib-His-D-2-Nal-D-Phe-Lys-NH₂
- **Aromatics:** D-2-Nal (large naphthyl chromophore, ε₂₈₀ ≈ 5000), D-Phe → strong 280 nm; the naphthyl dominates
- **HPLC column:** C18, short run; retention ~4–6 min on a 10-min gradient
- **MS adducts:** [M+H]⁺ = 712.4
- **Key impurities:** D/L-2-Nal stereoisomer (racemization; identical mass, different retention); D-Phe→L-Phe; Aib→Ala; His oxidation
- **Gray-market failure modes:** high purity on Finnrick (avg score > 6); most common issue is under-dose on CJC/Ipa blends. Because ipamorelin is a small pentapeptide with distinctive D-2-Nal, HPLC-UV is genuinely decisive for identity
- **Finnrick test count:** 273 samples / 51 vendors

### 4.8 GHK-Cu (copper tripeptide-1)

- **Formula / MW:** C₁₄H₂₄N₆O₄ (peptide) + Cu²⁺; GHK-Cu complex: C₁₄H₂₂CuN₆O₄ / 402 Da (complex); 340.38 Da (peptide only)
- **Sequence:** Gly-His-Lys-Cu²⁺
- **Aromatics:** His only (weak absorber ~210 nm); no true aromatics for 280 nm → use 214 nm for peptide quantification
- **HPLC column:** HILIC or ion-pair RP; free GHK is highly polar and pokes off a C18 within 1–2 min of dead time. Ion-pair with TFA is marginal; HILIC or a mixed-mode column (e.g., Primesep) is preferred. MZ Biolabs lists HILIC + Amide columns in their disclosed chemistry, which fits GHK-Cu.
- **MS adducts:** [M+H]⁺ = 341.2 (peptide), [M+H]⁺ = 403.1 (Cu complex). Copper isotope pattern (⁶³Cu / ⁶⁵Cu ≈ 69/31) is a diagnostic signature in MS.
- **Copper quantification:** the *active* is the copper complex. Testing GHK-Cu potency therefore requires **ICP-MS copper quantification** in addition to peptide content. A vial of "GHK-Cu 100 mg" should deliver both ~240 mg of tripeptide-peptide-content and the stoichiometric Cu; gray-market vendors sometimes sell GHK (free) or GHK acetate mixed with CuSO₄ that doesn't actually chelate.
- **Key impurities:** free GHK (unchelated), free Cu²⁺, Cu(OH)₂ precipitate, 2:1 and 1:2 GHK:Cu stoichiometry mix
- **Gray-market failure modes:** copper stoichiometry deviation; vendors claiming "GHK-Cu" selling GHK only
- **Finnrick test count:** 416 samples / 54 vendors

### 4.9 Melanotan-II

- **Formula / MW:** C₅₀H₆₉N₁₅O₉ / 1024.18 Da
- **Sequence:** Ac-Nle-cyclo[Asp-His-D-Phe-Arg-Trp-Lys]-NH₂ (cyclic lactam between Asp and Lys)
- **Aromatics:** Trp, D-Phe, His → strong 280 nm (Trp dominant)
- **HPLC column:** C18, 10–40% B over 15 min
- **MS adducts:** [M+H]⁺ = 1025.2, [M+2H]²⁺ = 513.1
- **Key impurities:** linear (uncyclized) MT-II (+18 Da water, open ring); racemized D-Phe; Trp oxidation
- **Gray-market failure modes:** MT-I substitution (linear, +117 Da; structurally very different but still α-MSH-active so user may not notice lower tanning response); incomplete cyclization is the biggest real quality issue

### 4.10 Tesamorelin

- **Formula / MW:** C₂₂₁H₃₆₆N₇₂O₆₇S / **5135.86 Da** (round-2 correction R2-M2; earlier draft said 5195.89 — off by 60 Da vs PubChem, Wikipedia, vendor datasheets)
- **Sequence (44-mer, trans-3-hexenoyl-Tyr¹ modification):** (hexenoyl)Y-A-D-A-I-F-T-N-S-Y-R-K-V-L-G-Q-L-S-A-R-K-L-L-Q-D-I-M-S-R-Q-Q-G-E-S-N-Q-E-R-G-A-R-A-R-L (canonical GHRH sequence)
- **Aromatics:** Tyr¹, Phe⁶, Tyr¹⁰ → moderate 280 nm. Contains Met²⁷ (oxidation hotspot)
- **HPLC column:** C18, 20–55% B over 20 min; typical retention 10–12 min
- **MS adducts (recomputed, R2-M2):** [M+4H]⁴⁺ = **1284.97**, [M+5H]⁵⁺ = **1028.18**
- **Key impurities:** Met²⁷-sulfoxide (+16); des-hexenoyl (hydrolysis of N-terminal modification, loss of 97 Da); deamidation at Asn⁸, Asn³⁵, Gln¹⁶, Gln²⁴, Gln³⁰, Gln³⁶ (many sites)
- **Finnrick test count:** 347 samples / 39 vendors

---

## 5. Specific failure findings from Finnrick's public data

### 5.1 Headline statistics (as of 2026-04-23)

**Claim:** Finnrick's public product catalog as of 2026-04-23 reports: Retatrutide 2,489 samples / 164 vendors; Tirzepatide 1,776 / 136; GHK-Cu 416 / 54; Tesamorelin 347 / 39; Semaglutide 309 / 51; Ipamorelin 273 / 51; CJC-1295 248 / 47; Melanotan II 140 / 34; Cagrilintide 106 / 18; PT-141 96 / 27; TB-500 83 / 11; BPC-157 501 / 72; Mazdutide 28 / 10; Survodutide 26 / 7; Sermorelin 23 / 5.
- **Confidence:** C1 (fetched from finnrick.com/products 2026-04-23)
- **Source:** https://www.finnrick.com/products
- **Date checked:** 2026-04-23

**Claim:** Finnrick's CEO states in the "What we believe in" post that "roughly one in three [peptide samples] fail basic expectations" — interpreted as a composite of purity, quantity accuracy, and batch information.
- **Confidence:** C1 (verbatim from blog)
- **Source:** https://www.finnrick.com/blog/what-we-believe-in
- **Date checked:** 2026-04-23
- **Notes:** Exact definition of "fail" is not given in the blog; Finnrick's scoring system maps ≥F (Fraud) and E (Bad) to outright failure, with D (Poor) arguably failing.

### 5.2 Endotoxin findings

**Claim:** Of 140 samples run through Finnrick's endotoxin add-on through January 2026: 63% no detectable endotoxins; 29% small quantities below limit-of-quantification (LoQ); 8% quantifiable above threshold.
- **Confidence:** C1
- **Source:** https://www.finnrick.com/blog/why-endotoxin-testing-matters-for-peptides
- **Date checked:** 2026-04-23
- **Notes:** Finnrick uses an internal threshold of 40 EU per vial (derived from FDA 5 EU/kg × 80 kg assumed body weight). LAL, rFC, or rCR assays are accepted; detection limit ~0.01 EU/mL.

### 5.3 Semaglutide "market check"

The Finnrick blog does not contain a single dedicated "semaglutide market check" post as of 2026-04-23. Specific vendor results are accessible via per-product pages — e.g., the Peptide Sciences / Semaglutide product page (public via https://www.finnrick.com/products/semaglutide/peptide-sciences) reports 15 samples from that vendor tested between 17 Dec 2024 and 1 Mar 2026, all scoring above 5, earning a confident B rating.

**Naming vendors with poor ratings** (Finnrick publishes publicly; cite their page, do not editorialize): vendor-level F and E ratings are listed in the **Corrections** and **Vendors** sections. Because the dataset is dynamic and ratings recalculate, any specific vendor called out as "F" today may move. Readers should cite the live finnrick.com URL + check-date rather than this document.

### 5.4 Sequence-level failures (LC-MS/MS findings)

**Claim:** Chromate Labs has been reported in community discussion to have issued corrected COAs in 2026 after Finnrick cross-referencing against MZ Biolabs or Krause flagged a discrepancy; specific incidents reported include a Coastal Peptides cagrilintide/semaglutide mix-up and a tirzepatide/retatrutide mix-up.
- **Confidence:** C4 (round-2 downgrade from C1; the specific incidents were not independently re-verifiable against finnrick.com/corrections or any primary source on 2026-04-23. The general pattern — UV-only peptide identification producing ambiguous IDs for similar-mass peptides — is well-supported on first principles).
- **Source:** Round-1 research doc `research/_sources/partner-labs-summary.md`; community forum reporting; not re-confirmed on Finnrick's Corrections page in round 2.
- **Date checked:** 2026-04-22 (round 1) / 2026-04-23 (rechecked; not re-verifiable)
- **Notes:** These would be **lab** errors, not vendor frauds. They illustrate why Finnrick's cross-partner consistency analysis exists. A lab running HPLC-UV without MS confirmation cannot reliably distinguish semaglutide from cagrilintide (similar size, similar retention). Once MS is applied, such errors are catchable.

### 5.5 Counterfeit / substitution cases

FDA public alerts (not Finnrick data, but relevant): FDA and FBI have warned about counterfeit compounded semaglutide products that contain no semaglutide or contain a different GLP-1 (liraglutide). These are typically caught by intact mass — liraglutide is 3751 Da vs semaglutide 4114 Da, a 363 Da shift that a single-quad LC-MS identifies immediately.

**Claim:** FDA has documented counterfeit compounded semaglutide and tirzepatide with false labeling including non-existent compounding pharmacies and unknown drug mixtures.
- **Confidence:** C1
- **Source:** https://www.fda.gov/drugs/postmarket-drug-safety-information-patients-and-providers/fdas-concerns-unapproved-glp-1-drugs-used-weight-loss
- **Date checked:** 2026-04-23

---

## 6. Cost implications

Peptide testing is structurally more expensive than small-molecule testing. Specific drivers:

### 6.1 MS is (usually) required

For a serious identity answer, you want at least intact MS and ideally MS/MS fragmentation. That means Finnrick routes to partners with MS capability (Krause, MZ Biolabs, Janoshik) for a non-trivial share of peptide tests. HRT testing, by contrast, can reasonably be done on HPLC-DAD alone.

### 6.2 Reference standards are 10–100× more expensive

| Analyte | Pharmacopoeial RS (per mg) | Research-grade bulk (per mg) |
|---|---|---|
| Estradiol valerate (USP) | ~$3–5 | ~$0.10 (API grade) |
| Testosterone cypionate (USP) | ~$3–5 | ~$0.10 |
| Semaglutide (USP impurity ARM; parent via Sigma CRM) | $500–2,000 | $5–50 (API grade, Bachem or China CRO) |
| Tirzepatide (Sigma/Cerilliant CRM) | $500–2,000 | $10–80 |
| Retatrutide (no USP RS) | N/A | $50–200 (research-grade only) |
| BPC-157 (no USP/EP RS) | N/A | $10–50 |

See `research/06-reference-standards/peptide-standards.md` for the full picture. The practical community-lab workflow is: buy a certified RS once, use it to characterize a bulk API, then use characterized bulk as working standard — acceptable methodology but only if the initial characterization is rigorous (MS + amino acid analysis + repeat-day injection consistency).

### 6.3 Narrower dynamic range at 214 nm

A steroid assay at 254 or 280 nm in a DAD can span 3–4 orders of magnitude of linear UV response. At 214 nm with TFA gradient, the usable linear range is typically 2 orders. That means: more dilutions, more injections, more solvent, more labor per sample.

### 6.4 Commercial prices observed

| Service | Test | Price | Turnaround |
|---|---|---|---|
| Janoshik | Anabolic oil (steroid) — basic | $120 | 1–8 days |
| Janoshik | GLP-1 peptide — HPLC-UV + MS | $300 | 1–8 days |
| Janoshik | HGH 22 kDa | $420 | 1–8 days |
| Janoshik | Full GLP-1 panel (purity + sterility + endotoxin + metals) | ~$828 | 1–8 days |
| Finnrick | Peptide potency + purity (free for listed 15) | $0 | ~4–8 weeks |
| Finnrick | Endotoxin add-on | ~$110 (per community reporting; not on public methodology page) | ~4–8 weeks |
| Simec AG (Swiss) | HPLC-UV quantitative steroid | ~$300 | ~2–4 weeks |
| Commercial contract peptide LC-MS method development (e.g., Krause, Catalent) | Full method dev + validation | $5,000–15,000 | weeks–months |

**Claim:** Janoshik Analytical lists basic anabolic steroid screen at $120 and GLP-1 peptide test at $300 with 1–8 day turnaround, worldwide shipping, crypto payment accepted.
- **Confidence:** C1
- **Source:** `research/_sources/janoshik-pricing-2026-04-22.md` (fetched 2026-04-22)
- **Date checked:** 2026-04-22

### 6.5 Implication for a home lab

- **Identity and purity by HPLC-UV at 214 nm:** feasible with $5–15k used HPLC + VWD/DAD (see `research/05-home-lab-setup/budget-tier-silver.md`), but only if you are disciplined about reference standards, mobile phase reproducibility, and column life. BPC-157 and TB-500 require you to skip 280 nm entirely.
- **Sequence confirmation:** outsource. The used-LC-MS/MS economics don't close for a hobby-scale operation.
- **Endotoxin:** outsource unless you commit to a dedicated LAL workstation (see `research/03-microbial-pyrogen/home-lab-endotoxin-feasibility.md`). Peptides-in-water (reconstituted) are actually an easier LAL matrix than steroid-in-oil.

The hybrid model that falls out of this: **do your own HPLC-UV potency check at home for a peptide you use regularly; send ~1 in 5 batches out to Finnrick (free) or Janoshik ($300) for independent confirmation.**

---

## 7. Community-lab peptide specialists

Who to send what, in practice. See `research/01-existing-services/partner-labs.md` for deeper lab-by-lab methodology.

### 7.1 MZ Biolabs — sequence-confirmation specialist

- **Location:** 2102 N Country Club Rd, Suite C, Tucson, AZ 85716. DEA licensed (RP0584676).
- **Instruments:** 3× Bruker Compact QTOF, Thermo LTQ Velos Pro, Waters Acquity + nanoAcquity UPLC. C8 / C18 / HILIC / Amide column chemistry disclosed.
- **Strength:** MS/MS sequence reading, impurity structural ID, HILIC for GHK-Cu / polar peptides.
- **Finnrick role:** orthogonal confirm partner. 178 tests for Finnrick (as of 2026-04-22). When Krause or another partner flags an anomaly, MZ is where the sample often goes for MS².
- **Sample access:** not obviously public-friendly — geared to research/medical clients. Contract for individual vial sequence ID is typically $200–300 per full panel per community-report triangulation.
- **Best for:** any peptide where identity is questioned; research peptides without USP RS; polar peptides like GHK-Cu needing HILIC.

### 7.2 Krause Analytical Labs — Finnrick's house HPLC lab

- **Location:** Austin, TX. Principal scientist Mark Krause, 45+ years experience.
- **Instruments:** HPLC-UV; LC-MS; GC-MS; ICP; UV; IR. USP monograph-style workflows.
- **Finnrick role:** largest partner, 3,836 tests through 20 Mar 2026 (~56% of total Finnrick test volume).
- **Strength:** high-throughput peptide HPLC-UV purity/potency; advertised LC-MS workflows for semaglutide, tirzepatide, BPC-157, CJC-1295.
- **Sample access:** commercial contract lab; pricing is private; direct-to-public submissions are not advertised. Access the methodology is via Finnrick — ship your vial to Finnrick for free test, it may end up at Krause.

### 7.3 Janoshik Analytical — the gray-market Schelling point

- **Location:** Prague, Czech Republic (Kaprova 42/14, Staré Město; registered 25 Oct 2022).
- **Founder:** "Peter Magic" (Slovak, former amateur weightlifter).
- **Instruments (disclosed):** HPLC-UV, LC-MS/MS, GC-MS, LAL, ICP-MS, sterility.
- **Pricing:** $300 GLP-1 peptide; $120 anabolic basic; $420 HGH 22-kDa; $828 full GLP-1 panel. Int'l shipping $50–75. Crypto payments accepted.
- **Turnaround:** 1–8 days (fastest of any community-serving lab).
- **Not ISO 17025.** COAs do not include raw data by default (paywalled).
- **Data breach note:** a 2026 data breach exposed customer shipping info; customers who cared about anonymity have reconsidered.
- **Best for:** cheap, fast, worldwide peptide testing. Gray-market Schelling point: if a vendor shows a "Janoshik COA," community buyers trust it more than any vendor-provided COA.

### 7.4 Others in Finnrick's partner network

- **BTLabs** (btlabtesting.com) — 1,446 tests; peptides, vitamins, nutraceuticals. 25+ years claimed. Location, pricing, instruments not publicly disclosed.
- **Chromate Labs** (chromate.org) — 660 tests; dietary-supplement flavor. Notable for two publicized early-2026 misidentifications (Cagri→Sema; Reta→Tirz), which were corrected after Finnrick triangulation.
- **TrustPointe Analytics** — 315 tests; Michigan-based; 3–5 day turnaround; Mots-C and NAD+ specialties advertised.
- **Freedom Diagnostics** (Franklin, TN) — 213 tests; research-use-only peptide purity; fast turnaround.

### 7.5 Simec AG (Swiss) and AnabolicLab — steroid, not peptide

Included for completeness: Simec AG is the Swiss HPLC-UV shop AnabolicLab redirects to. Not a peptide testing destination — they're the HRT/anabolic analog of MZ Biolabs. Note if you need HRT testing, not peptide testing.

---

## 8. Decision framework

Given all the above, practical advice for someone self-sourcing peptides:

1. **For any GLP-1 (semaglutide / tirzepatide / retatrutide):** ship a vial to Finnrick (free, 4–8 weeks, US only). If you need faster or non-US, use Janoshik ($300 + shipping).
2. **For BPC-157 / TB-500 / GHK-Cu (peptides with no USP RS and tricky UV detection):** Finnrick still works but prioritize vendors Finnrick has already tested — their partner labs have validated 214 nm methods for these peptides. Trust a Janoshik COA more than an in-house vendor COA.
3. **For a peptide not in any lab's standard catalog** (e.g., a newly-published research peptide): expect to pay for method development at $5k+ at a contract lab, or accept that you can get rough HPLC purity but not assured identity.
4. **For repeat verification of a vendor you trust:** a home HPLC-UV setup with 214 nm VWD can do ongoing potency checks on reconstituted material; send 1 in 5 batches out for MS confirmation.
5. **For sterility / endotoxin:** LAL gel-clot kit at home is legitimately feasible for reconstituted peptides (aqueous matrix, unlike HRT oils which break LAL). See `research/03-microbial-pyrogen/home-lab-endotoxin-feasibility.md`.

---

## 9. Gaps and things this document does not cover

- **Aggregate-state characterization.** Peptide aggregation / fibrillation is a sterility-relevant but rarely-tested property. Dynamic light scattering (DLS) or SEC-HPLC catches it; neither is in any community-lab standard panel.
- **Chiral purity.** D/L stereochemistry on individual residues is not caught by intact MS, not easily by HPLC-UV, and only detectable by chiral HPLC or Marfey's derivatization + LC-MS. No community lab runs this routinely.
- **Host-cell protein and DNA impurities.** For recombinant peptides (some knockoff "semaglutide" is reportedly recombinant from E. coli or yeast rather than SPPS), HCP and residual DNA must be measured per ICH Q6B. No community lab runs this.
- **Per-vendor failure rate by peptide.** Extracting vendor-specific failure rates from Finnrick is possible but out of scope for this doc — their product pages have per-vendor ratings that change over time; any static summary would be stale within weeks. The `research/01-existing-services/finnrick.md` doc has the methodology pointer.
- **Retatrutide reference standard sourcing in detail.** Bachem and CS Bio both sell research-grade retatrutide as of 2026; specific pricing requires a quote and varies by volume.

---

## Sources

### Finnrick and community labs
- https://www.finnrick.com/ — retrieved 2026-04-23
- https://www.finnrick.com/about/testing-methodology — retrieved 2026-04-22
- https://www.finnrick.com/labs — retrieved 2026-04-22
- https://www.finnrick.com/products — retrieved 2026-04-23 (sample counts per peptide)
- https://www.finnrick.com/blog/why-endotoxin-testing-matters-for-peptides — retrieved 2026-04-23
- https://www.finnrick.com/blog/what-we-believe-in — retrieved 2026-04-23
- https://www.finnrick.com/blog/how-to-read-a-certificate-of-analysis-coa — retrieved 2026-04-23
- https://www.finnrick.com/products/semaglutide/peptide-sciences — retrieved 2026-04-23
- https://www.mzbiolabs.com/mzbiolabs/our-techniques/ — retrieved 2026-04-23
- https://www.krauselabs.com/ — retrieved 2026-04-22 (round 1)
- https://janoshik.com/ — retrieved 2026-04-22 (round 1)

### Peptide chemistry references
- https://pubchem.ncbi.nlm.nih.gov/compound/Semaglutide (C₁₈₇H₂₉₁N₄₅O₅₉) — retrieved 2026-04-23
- https://pubchem.ncbi.nlm.nih.gov/compound/Tirzepatide (C₂₂₅H₃₄₈N₄₈O₆₈) — retrieved 2026-04-23
- https://pubchem.ncbi.nlm.nih.gov/compound/Retatrutide (C₂₂₁H₃₄₂N₄₆O₆₈) — retrieved 2026-04-23
- https://pubchem.ncbi.nlm.nih.gov/compound/Bpc-157 (C₆₂H₉₈N₁₆O₂₂) — retrieved 2026-04-23
- https://pubchem.ncbi.nlm.nih.gov/compound/Tesamorelin — retrieved 2026-04-23
- https://pubchem.ncbi.nlm.nih.gov/compound/Cjc-1295 — retrieved 2026-04-23
- https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?tab=structure&ligandId=11429 (tirzepatide sequence) — retrieved 2026-04-23
- https://www.guidetopharmacology.org/GRAC/LigandDisplayForward?tab=structure&ligandId=13769 (retatrutide sequence) — retrieved 2026-04-23
- https://en.wikipedia.org/wiki/Semaglutide — retrieved 2026-04-23
- https://en.wikipedia.org/wiki/Tirzepatide — retrieved 2026-04-23
- https://en.wikipedia.org/wiki/Retatrutide — retrieved 2026-04-23
- https://en.wikipedia.org/wiki/BPC-157 — retrieved 2026-04-23
- https://en.wikipedia.org/wiki/Ipamorelin — retrieved 2026-04-23
- https://en.wikipedia.org/wiki/TB-500 — retrieved 2026-04-23
- https://en.wikipedia.org/wiki/CJC-1295 — retrieved 2026-04-23
- https://en.wikipedia.org/wiki/Copper_peptide_GHK-Cu — retrieved 2026-04-23
- https://pubmed.ncbi.nlm.nih.gov/21548867/ (BPC-157 pentadecapeptide reference) — retrieved 2026-04-23

### Regulatory
- https://www.fda.gov/drugs/postmarket-drug-safety-information-patients-and-providers/fdas-concerns-unapproved-glp-1-drugs-used-weight-loss — retrieved 2026-04-23
- https://www.usp.org/biologics/peptides/glp1s — retrieved 2026-04-22 (round 1)

### Internal cross-references
- `research/01-existing-services/finnrick.md`
- `research/01-existing-services/partner-labs.md`
- `research/02-analytical-chemistry/hplc.md`
- `research/02-analytical-chemistry/lc-ms.md`
- `research/03-microbial-pyrogen/home-lab-endotoxin-feasibility.md`
- `research/05-home-lab-setup/budget-tier-silver.md`
- `research/06-reference-standards/peptide-standards.md`
- `research/_sources/finnrick-blog.md`
- `research/_sources/finnrick-labs.md`
- `research/_sources/finnrick-testing-methodology.md`
- `research/_sources/janoshik-pricing-2026-04-22.md`
- `research/_sources/partner-labs-summary.md`