What Raw Analytical Data Adds to a Short Certificate
Identify the question behind the request
“Please send the raw data” can mean very different things to a supplier and a laboratory. One party may send a screenshot; another may offer instrument files that require specialist software. Begin with the uncertainty you want to resolve: peak assignment, signal processing, a calculation or the connection between a sample and a result.
Ask what files exist and which are original observations, processed outputs or summaries. A PDF chromatogram may be useful, but its appearance does not tell you whether processing settings or acquisition metadata have been retained elsewhere.
Look beyond the picture
A chromatographic trace needs context: sample reference, analytical method, detector channel, relevant axes and the integration used for the stated calculation. A mass spectrum likewise needs an explanation of what the displayed signal represents and how it supports the reported assignment. These are review questions, not a recipe for reproducing a method.
The laboratory-record section of ICH Q7 discusses underlying analytical records. Its pharmaceutical setting should be recognised. A research buyer can still ask which evidence supports the particular summary supplied, without assuming every file must be publicly downloadable.
Keep processing history in view
A result may change when data are reprocessed. Ask whether the supplied output is the version used for the certificate and whether relevant changes have been documented. A disagreement between two PDFs may reflect a revision, different processing or a different sample; the images alone may not resolve it.
The OECD data-integrity document provides a primary reference for considering data through their life cycle. Preserve the explanatory correspondence with the analytical files so that later reviewers can understand why a particular output was accepted.
Make a proportionate review
For an initial screen, a complete report and intelligible supporting plots may answer the question. A contested result or method transfer may require deeper discussion with the laboratory, including access to native files and appropriately qualified review. More files do not automatically create a better conclusion.
Start with the available material in the COA library and record what remains unavailable. Do not infer unreported identity, content or microbiological properties from a single attractive trace.
Finish by writing what the added evidence established, what it did not establish and whether the remaining uncertainty matters for the planned study. That short assessment makes the request for raw data useful rather than merely accumulating attachments.
Related research reading
Explore the research blog and original batch-document library for further context.