I see a lot of people posting COAs without understanding what they're looking at. This guide will teach you to read a Certificate of Analysis for compounded or research-grade peptides like semaglutide. Bookmarkable reference. What is a COA? A Certificate of Analysis is a document from a testing laboratory reporting the results of quality testing on a specific batch of product. It tells you what's in the vial, how pure it is, and whether it meets specifications. Key sections of a peptide COA: 1. Identity — "Is this actually semaglutide?" - Method: Usually mass spectrometry (ESI-MS or MALDI-TOF) - What to look for: Molecular weight should match semaglutide (MW: 4113.58 Da). The observed mass should be within ±2 Da of theoretical. - Red flag: Mass doesn't match, or identity testing wasn't performed. 2. Purity — "How clean is it?" - Method: HPLC (High Performance Liquid Chromatography) or UPLC - What to look for: Purity ≥ 95% is acceptable. ≥ 97% is good. ≥ 99% is excellent. - How it works: The sample is dissolved and pumped through a column. Different molecules elute at different times. The main peak (semaglutide) is expressed as a percentage of total peak area. - Red flag: Purity below 95%, large unidentified peaks, or no chromatogram provided (just a number without supporting data). 3. Content/Potency — "How much is actually in the vial?" - Method: Quantitative HPLC against a reference standard - What to look for: The measured amount vs. label claim. A 5mg vial should contain 4.5-5.5mg (90-110% of label). Ideally 95-105%. - Red flag: Content below 90% or above 115% of label claim. Under-filling is quality control failure. Over-filling beyond 110% suggests poor manufacturing controls. 4. Sterility (if tested) - Method: USP <71> sterility testing (14-day incubation in growth media) - What to look for: "No growth observed" / "Passes USP <71>" - Red flag: Not tested, or growth detected. NOTE: Most research-grade COAs don't include sterility testing. Only 503B pharmacies are required to test. 5. Endotoxin (if tested) - Method: LAL (Limulus Amebocyte Lysate) test - What to look for: Results should be below 5 EU/kg for injectable products. Often reported as <0.25 EU/mL. - Red flag: Not tested (for injectables) or above threshold. 6. Appearance / Solubility - What to look for: White to off-white powder (lyophilized). Clear and colorless solution (reconstituted/liquid). - Red flag: Yellow discoloration, visible particulates. I'll do a walkthrough of an actual COA in the next post.

Walkthrough of a real COA (anonymized): Here's what a Janoshik COA for semaglutide typically looks like. I'm using data from a recent community-submitted test: ``` CERTIFICATE OF ANALYSIS Lab: Janoshik Analytical Sample ID: JS-2026-01-4872 Date Received: January 14, 2026 Date Reported: January 21, 2026 Sample Description: White lyophilized powder, labeled "Semaglutide 5mg" TEST METHOD RESULT SPECIFICATION Identity ESI-MS Confirmed MW 4113.58 ± 2 Da (Observed: 4113.4 Da) Purity HPLC-UV 98.3% ≥ 95.0% (220nm) Content HPLC vs. std 4.87 mg 4.50 - 5.50 mg (97.4% of label) Residual Solvents GC-FID < LOQ Meets USP <467> Water Content Karl Fischer 1.2% ≤ 5.0% CONCLUSION: Sample meets all tested specifications. ``` How to interpret this: - Identity confirmed at 4113.4 Da — that's within 0.2 Da of theoretical. This is semaglutide. - 98.3% purity — excellent. The remaining 1.7% is likely minor synthesis impurities (truncated sequences, oxidized variants). Nothing unusual. - 4.87mg content — 97.4% of the labeled 5mg. You're getting slightly less than advertised but well within acceptable range. - Residual solvents below limit of quantitation — good. Means the lyophilization was thorough. - 1.2% water content — normal for lyophilized peptide. Up to 5% is acceptable. This is a solid COA. The product is what it claims to be, at the right purity and potency.

This is incredibly helpful. A few clarifying questions: 1. When the purity says 98.3%, what exactly is the other 1.7%? Is it dangerous? 2. How do I know a COA is real and not fabricated? I've heard vendors can just make fake COAs in Photoshop. 3. What's the difference between a vendor-provided COA (from the manufacturer) and a third-party COA (from Janoshik, etc.)? Which should I trust more?

Let me tackle these: 1. The "other 1.7%": In peptide synthesis, common impurities include: - Deletion sequences — the peptide chain missing one or more amino acids - Oxidized variants — methionine residues oxidized during synthesis or handling - Truncated sequences — synthesis stopped before completion - Deamidation products — asparagine converting to aspartate - TFA (trifluoroacetic acid) — residual counterion from purification - Acetic acid salt forms — if the peptide was acetate-exchanged At 98.3% purity, the 1.7% impurities are at very low levels. For injectable peptides, the concern is whether any single impurity exceeds 0.5% and whether it's been identified. Reputable labs will flag any individual impurity above 0.5% on the chromatogram. At sub-1% levels each, these synthesis byproducts are generally considered safe for research purposes (and pharmaceutical-grade GLP-1 products have similar impurity profiles — they're just more thoroughly characterized). 2. Fake COAs: This is a real problem. Ways to verify: - Contact the lab directly. Janoshik has a verification page where you can enter the sample ID to confirm it's genuine. Use it. - Check for consistency. A fake COA often has mismatched fonts, wrong test methods, or results that are suspiciously "perfect" (e.g., exactly 100.0% purity — real analytical data has decimal variation). - Look for the chromatogram. A real COA usually includes the actual HPLC chromatogram image. Faking a realistic chromatogram is much harder than faking a table of numbers. - Batch-specific data. A real COA has a unique sample ID and batch number. If a vendor uses the same COA for every batch, it's likely fake. 3. Vendor vs. third-party COA: Always prioritize third-party testing. A vendor-provided COA is the manufacturer grading their own homework. It may be legitimate, but there's an inherent conflict of interest. Third-party labs like Janoshik, Finnrick, or Vanta have no financial stake in the product passing. That independence is what makes their results credible. Best case scenario: the vendor provides their own COA AND a community member independently tests the same batch at a third-party lab, and the results agree. That's the gold standard of verification.

Great post Annie. I want to add some detail on reading HPLC chromatograms since the OP mentioned them: What you're looking at: An HPLC chromatogram is a graph with time (minutes) on the X-axis and detector response (absorbance units or millivolts) on the Y-axis. Each "peak" represents a compound eluting from the column. For semaglutide, expect: - Main peak at a retention time typically between 12-18 minutes (depends on the method). This is the semaglutide. - Small peaks before the main peak — these are more polar impurities (hydrophilic, elute faster) - Small peaks after the main peak — these are less polar impurities or oxidized forms - Solvent front at ~2-3 minutes — ignore this, it's just the injection solvent What "98.3% purity by HPLC" means: The area under the main semaglutide peak represents 98.3% of the total area of ALL peaks. This is "area percent" or "area normalization" purity. Important caveat: HPLC purity depends heavily on the method. A method that uses a short column, fast gradient, or UV detection at 280nm might miss impurities that would show up with a longer column, slower gradient, or detection at 220nm. This is why comparing purity numbers across different labs requires knowing the method details. A "99.5% pure" result from one lab is not necessarily better than "97.8% pure" from another if they're using different methods. The lab using the more sensitive method might simply be detecting more impurities that the other lab missed.