Batch Identity Testing: The Missing Half of Many Peptide Purity Conversations
By Marcus Reid — Mon Sep 28 2026
Batch Identity Testing: The Missing Half of Many Peptide Purity Conversations — my honest, first-person take, backed by data from the 289 peptide vendors I track. Research use only.
I used to treat a COA like an answer — until a batch arrived with the right purity number and the wrong molecular weight. After that, I stopped outsourcing my skepticism. Batch identity testing changed how I pick vendors, interpret LC-MS traces, and even which peptides I trust in an experiment.
Why batch identity testing deserves equal billing to “purity”
When people talk about peptide purity they usually mean a percent on a COA: 95% here, 98% there. I understand why — percent purity is concise and easy to compare. But purity is a downstream summary. It doesn’t tell you whether the main peak is the peptide you ordered, a truncated analogue, or a solvent adduct. That’s where peptide identity testing comes in: it closes the loop between “this vial looks pure” and “this vial is actually the sequence and mass I expect.”
In my experience, skipping identity testing is how labs accidentally run months of work on the wrong molecule.
What I see across vendors (data from my tracking)
In my vendor database I track 289 vendor profiles. Only 65 — about 22% — publish named-lab COAs for individual lots. That matters: a vendor COA from an unnamed in-house instrument is useful, but a named third-party LC-MS report is far easier to verify independently. As of now, 1 currently have published editorial assessments, with an average assessed rating of 4.70/5; 1 assessed vendors clear a 4.5/5 rating. Those numbers tell me two things: many vendors still treat batch transparency as optional, and a few well-documented vendors consistently stand out when they tie purity figures to clear LC-MS evidence.
If you’re ordering for sensitive assays, assume a COA percent alone is insufficient unless you can see lot-level LC-MS evidence and confirm the identity yourself or via a named lab.
My go-to checklist: the BATCH-ID 4-step check
I use a small, repeatable framework I call BATCH‑ID. It’s short, practical, and I run it before I accept a lot into the freezer.
1. Batch COA match — Does the COA explicitly name the lot number printed on the vial? 2. Authentic LC‑MS — Is there a named-lab LC-MS peptide COA showing the expected monoisotopic mass (or deconvoluted mass) and retention behavior? 3. Sequence confirmation — Is there MS/MS or orthogonal evidence (amino-acid analysis, Edman for short peptides, or peptide mapping) for higher-risk sequences? 4. Trace review — Do the chromatogram and mass spectrum show a dominant, correctly assigned peak and no suspicious major impurities?
I use this checklist every time I switch vendors or accept a new lot. It only takes a few minutes to reject a lot that fails two items.
Reading an lc-ms peptide coa without a PhD (practical tips)
I don’t rely on the vendor’s summary sentence; I look at the picture.
- Check the mass: the deconvoluted or monoisotopic mass should match your theoretical mass within instrument error (typically ±0.01–0.1 Da for high-res). - Inspect the isotopic envelope: a correct peptide peak has a characteristic envelope. Missing isotopic structure or a single strange m/z cluster is a red flag. - Look for adducts and truncations: common nearby peaks are [-H2O], sodium adducts, and truncated sequences (M-? residues). If one of those is your “main” peak, that’s not a 95% correct peptide. - Verify retention time consistency: if the vendor provides a retention time, note it. The same peptide prepared on your system should elute in a comparable window (allowing for column/gradient differences). - Prefer named labs: a named-lab LC-MS peptide COA often includes method details — column, gradient, detector, resolution — which makes the result verifiable.
Below is a short table I use mentally when scanning a COA:
| COA element | What I expect | Immediate red flag | |---|---:|---| | Lot number | Matches vial | COA lacks lot or mismatch | | Mass report | Theoretical mass ± instrument error | Major mass shift | | Chromatogram | Single dominant peak | Multiple equal peaks | | Lab ID | Named third party | “In-house” only |
A counter-angle I sometimes argue in the room
The community mantra I hear is “insist on MS/MS for every single peptide.” I push back — carefully. For short, well-characterized peptides (under ~10 amino acids) with a clean intact mass, a high-quality LC-MS trace and an orthogonal check (amino acid analysis or HPLC retention profile) can be sufficient for routine experiments. Requiring MS/MS for every batch will delay projects, increase cost, and sometimes produce noisy spectra that are harder to interpret than a clean intact mass.
That said: if you’re working with novel sequences, post-translationally modified peptides, or anything where a single residue flip ruins the biology, MS/MS or sequence-level confirmation is non-negotiable. Use judgment: not every peptide needs the heaviest artillery, but every critical peptide needs identity evidence beyond a percent.
How to incorporate peptide batch testing into lab workflow
- Ask for the named-lab LC-MS peptide COA when you order. If the vendor won’t or can’t provide it for that lot, send the vial back or insist on a replacement. - Keep a COA log keyed to lot numbers and instrument traces. I keep mine beside chromatograms for quick cross-checking before an experiment. - Consider spot-checking: when budgets are tight, prioritize identity testing for peptides used in key assays, controls, or expensive experiments. - Train your team to spot obvious red flags (mass mismatch, multiple similar peaks, missing lot numbers). It’s cheap insurance.
For a list of peptides I commonly screen for identity issues, see /peptides-list. If you’re comparing suppliers, my vendor tracking lives at /vendors. And when you need to double-check theoretical masses quickly, I use /peptide-calculator before I even open a COA.
Final, blunt take
Purity percentages are necessary, but not sufficient. In my work, peptide identity testing — preferably via a named-lab LC-MS peptide COA and a quick BATCH‑ID check — has prevented more failed experiments than any single other step. Vendors that provide transparent, lot-level evidence earn my trust; those that don’t are on thin ice, no matter how pretty their percentage claims look.
*This article is for educational and research-use-only purposes. I am not a doctor. None of this content should be interpreted as guidance for human use.*
Frequently asked questions
What do you mean by "batch identity testing" and how is it different from peptide purity testing?
I use "batch identity testing" to mean the analytical checks that confirm a peptide batch actually contains the sequence and molecular species it is labelled as — in short, that the product is what you think it is. Purity testing typically reports how much of the sample is your target peptide versus other material (percentage purity), but it doesn’t always tell you whether the target peak is the correct sequence or whether similar masses or co-eluting species are present. In my experience, identity testing and purity testing are two sides of the same coin: purity answers “how much,” identity answers “which one.” Together they give a much clearer picture of sample quality. for educational and research-use-only purposes; this is not medical advice and no content should be treated as guidance for human use.
If my certificate of analysis shows high purity but identity testing flags a mismatch, what should I do?
I’ve seen this exact situation before — a high reported purity can mask a serious problem if the dominant material isn’t the intended sequence. When identity and purity disagree, I first treat the lot as suspect: document the discrepancy, stop using the material for any downstream experiments that rely on correct sequence, and contact the supplier with the data. In parallel I recommend obtaining orthogonal confirmation from an independent lab or using a different analytical approach to confirm the finding. The goal is to determine whether the issue is a labeling/supply-chain error, a degradation/truncation problem, or an analytical artifact. Resolve the root cause before proceeding. for educational and research-use-only purposes; this is not medical advice and no content should be treated as guidance for human use.
What are the common limitations of batch identity testing I should keep in mind?
From my perspective, identity testing is indispensable but not infallible. Limitations include method-dependent blind spots (some approaches miss low-level variants or certain modifications), dependence on quality reference standards, and the possibility of ambiguous results for very similar sequences or isobaric species. Sample handling and documentation also matter: a clean analytical result only reflects the analyzed aliquot, not necessarily the whole manufactured lot. Because of these limits, I always advocate for a layered approach — reliable reference material, orthogonal methods when results are critical, and clear traceability — rather than relying on a single test as the definitive answer. for educational and research-use-only purposes; this is not medical advice and no content should be treated as guidance for human use.
References
About the author
Marcus Reid: Marcus Reid spent a decade in software engineering before going deep into peptide research, product documentation, and the clinical literature. He writes about what the data and the paperwork actually say. He is not a doctor; PeptideTally content is educational and does not constitute medical advice.