Is it possible to accredit an untargeted fingerprinting method for food authenticity covering both the chemical measurements, database comparison and the final result?
I have been told that comparing a fingerprint to a database and the subsequent result cannot be cover by an accreditation. If you have any experience or examples in this issue, please let me know.
Of course. Accreditation to a laboratory SOP is common under ISO 17025. Just make sure you actually follow the laboratory SOP.
If the result of the accredited method (not only numerical but also the assessment) is repeatable and reproducible, following always the accredited procedure, then yes. This means that a database of non-targeted fingerprints should be validated, and tested continuously with blind samples. PTs are also needed. It is a difficult and complicated case.
In the absence of available PT standards PTs are not required. In this case no PT would be available and thus none would be required. You would need to be able to demonstrate method validation (running known fingerprints as "unknowns", method robustness, etc).