Identity is a separate question from purity
Purity asks how much of the vial is one thing. Identity asks whether that thing is the thing on the label. A lot can be 99 percent pure and 100 percent wrong.
The failure is not usually fraud. It is a mass spectrum accepted on its own. A mass tells you the atoms are present in the right count; it does not tell you they are arranged in the right order, and for a peptide the arrangement is the molecule. A single transposed pair of residues leaves the mass untouched.
So every entry on the register carries at least two identity determinations by different physical principles, and both appear on the certificate. For a peptide that is a mass measurement plus a peptide map. For a small molecule it is a chromatographic comparison against a reference plus an infrared spectrum.
What is on a certificate of analysis
| Line | What it has to say | Why a thin certificate omits it |
|---|---|---|
| Attribute | The property being measured, named the same way every lot. | Renaming the attribute between lots hides a specification that moved. |
| Method | The analytical method, by name, with its validation state. | Without a method the result is a number with no provenance. |
| Specification | The floor or the range the lot had to reach. | A result with no specification cannot fail, which is why it is left off. |
| Result | The measured value for this lot, not a typical value. | A typical value is a marketing figure wearing a laboratory coat. |
| Lot number | The same lot number printed on the vial and on the carton. | A certificate that does not tie to the container proves nothing about it. |
| Dates | Manufacture, release, and the retest date the storage data supports. | An open ended date lets old material stay young on paper. |
| Signature | A named release signature, held on file. | An unsigned certificate has nobody standing behind it. |
Lot traceability
Every lot has a genealogy, held as document GB‑140: the raw material lot it was made from, the supplier qualification in force on that date, the container closure lot, the fill date, the equipment used, the analyst, and the release signature.
The point of a genealogy is not tidiness. It is that a question arriving eighteen months late has a route to an answer. If a container closure lot turns out to be a problem, the genealogy names every lot filled into it, in minutes, without anyone reconstructing a memory of the week.
| Recorded | Held for |
|---|---|
| Raw material lot and its incoming test results | 24 months |
| Supplier qualification in force at that date | 24 months |
| Container closure lot | 24 months |
| Fill and finish record | 24 months |
| Release testing data, raw and reduced | 24 months |
| Retained sample of the filled lot | 24 months |
| Release signature | 24 months |
How a lot is released
This is the sequence. Nothing skips a step because a customer is waiting, and the only person who can release a lot is not the person who made it.
Incoming material is tested, not trusted
A supplier certificate arrives with the drum. It is filed, and then the material is tested against our own identity and assay methods as though the certificate did not exist. A mismatch stops the material at the door.
Assay
Quarantine hold
The batch record is written before the batch
Quantities, equipment, sequence and in process checks are fixed in advance. A record written afterwards is a description of what happened, not a control on it.
Signed as it runs
Never reconstructed
Fill and finish against the record
Filled under the conditions the record specifies, into a qualified container closure, with in process checks recorded as they happen rather than at the end.
In‑process checks
Recorded live
Release testing, every attribute
The full specification on the entry sheet. Not a subset, not the fast lines only. Sterility takes fourteen days and the lot waits for it.
No subset
14 day sterility
The data is reviewed against the specification
By a second person, line by line, with the raw data alongside the reduced result. An out of specification result opens an investigation, and an investigation has to close with a cause before anything else moves.
Raw beside reduced
Cause before close
Release, or not
A named signature releases the lot. The person who signs did not make it. A lot that misses a single line is not released, and no amount of retesting turns a failure into a pass without an investigation that explains the first result.
Not the maker
One line fails it
The lot enters the register with its documents
Certificate, identity package, method summary, stability summary and genealogy are linked to the lot number before the lot is available to an account.
GB‑110 · 118
GB‑140
What happens when a lot fails
A failed lot is quarantined and does not become a discounted lot, a research grade lot or a lot for a different market. There is one grade here and material that does not reach it does not leave.
Retesting until a result passes is not an investigation. A failure closes when there is a cause and a corrective action, or it closes as unexplained and the lot is destroyed. Both outcomes are recorded.
If a failure affects an entry rather than a lot, the entry is marked on revision on sheet G‑02 and stays marked until it is resolved. Two entries carry that mark today.
What is not claimed on this sheet
Grade Bio does not claim a facility registration, an establishment license, an accreditation or a certification it does not hold. Nothing on this site should be read as one, and if a line here ever reads like one, it is a defect and we want to hear about it.
What is described above is the release sequence the company runs. It is a process claim, and process claims are the kind that can be shown to a person who asks. The documents in table 4.1 of the entry sheets are how they get shown.
Grade Bio is a brand of Mountain Brands LLC. Entity formation is pending and no registration has been filed. Correspondence to jv@gradebio.com or sw@gradebio.com.
Pre‑launch document Entity pending