How to Read a Peptide Certificate of Analysis

A COA reports the results of defined tests performed on a defined lot. Its value depends entirely on which tests were run, by what method, and whether the document actually corresponds to the material in front of you.

It is not a safety assessment, not a statement of fitness for any use, and not evidence that the compound does anything. A flawless COA on an unapproved research compound tells you the chemistry was measured, and nothing whatsoever about biology.

Start here, because everything else is conditional on it. The lot number on the certificate must match the lot on the vial. A certificate for a different batch of the same product describes material you do not have.

A certificate with no lot number, or one supplied as a generic product-level document, cannot support any batch-specific claim. This is the most common gap and the easiest to check.

Identity is usually established by mass spectrometry against the expected mass. This is necessary but considerably weaker than it appears, because mass is not unique. Isomers share a mass exactly. A deletion sequence differs by one residue and may be close enough to pass a loose tolerance. Closely related analogues can be near-identical.

Stronger identity evidence comes from tandem mass spectrometry that sequences the peptide, from peptide mapping after enzymatic digestion, or from co-elution with a characterised reference standard. If a certificate states only an observed mass matching a calculated mass, it has established that the material is consistent with the claimed compound, not that it is that compound.

Purity is almost always reported as RP-HPLC area percent: the target peak's area as a fraction of total detected peak area. Two limitations follow directly from that definition and are rarely stated.

First, it only counts what the detector sees. UV detection at 214 nm responds to the amide bond and so detects peptides well, but a non-peptide impurity with no absorbance there — a residual salt, a scavenger, some solvents — contributes nothing to the denominator and is invisible. A 99 percent area figure is 99 percent of what was detected.

Second, anything that co-elutes with the main peak is counted as product. Deletion sequences are the usual offenders, because they are structurally similar by construction. A single gradient on one column cannot exclude them; an orthogonal method, with a different column chemistry or a different pH, is what gives confidence.

A lyophilised peptide is not pure peptide by mass. It contains water and counter-ions — commonly acetate or trifluoroacetate from purification — which can account for a substantial fraction of the powder. Peptide content, determined by amino-acid analysis or nitrogen determination, tells you how much of the mass is actually peptide.

A material can be 98 percent pure by HPLC and still be well under 80 percent peptide by mass. Both numbers are true and they answer different questions. Weighing out by vial mass without knowing peptide content introduces an error that no downstream analysis will reveal.

The counter-ion matters beyond mass accounting. Trifluoroacetate is biologically active in cell culture at concentrations that can appear in a TFA-salt peptide, which is why acetate salts are preferred for some work. A certificate that does not say which salt form was supplied has omitted information that can change an experiment's result.

This is also why a stated molecular mass can legitimately differ between suppliers for the same compound: one is quoting the free base and another the salt.

In-house testing by the manufacturer is normal and not in itself a concern. Third-party testing by an accredited laboratory is stronger, and accreditation to ISO/IEC 17025 means a competence assessment has taken place rather than merely a claim of one.

Check that the certificate names the testing party, carries a date, and that the date is consistent with the stated manufacture and retest dates. A certificate with no date supports nothing, because peptide purity changes with storage.

Running through these takes under a minute and catches most of what goes wrong.