Practical reference
Where to send samples. What to ask the lab. How to verify the result.
The vectors documented on this site are not invisible to commercial testing. They are invisible to routine regulatory testing. Independent ISO-accredited laboratories with validated mass-spectrometry methods can detect grayanotoxin, pyrrolizidine alkaloids, lupinine, and tremetol in honey, water, milk, and animal-product samples. This page is a practical reference for sending samples and interpreting the results.
Laboratories
| Laboratory | Location | Notes |
|---|---|---|
| Eurofins | National (Texas presence) | Largest global food-testing network. Most likely to have validated LC-MS/MS methods for grayanotoxins and pyrrolizidine alkaloids. Top recommendation for specific neurotoxin compounds. |
| AGT Labs | Houston, TX | ISO 17025 accredited; 30+ years; chemical testing including pesticides, heavy metals, food additives. Good general-screening lab. |
| A&B Labs | Houston, TX + Baton Rouge, LA | Full-service environmental and food testing. Useful for Texas/Louisiana regional sampling. |
| AFL (Analytical Food Laboratories) | Texas | 30+ years; full-service chemistry and microbiology; consultative engagement available. |
| Quanta Lab | San Antonio, TX | ISO-accredited; specifically offers mycotoxin testing; South Texas coverage. |
| UC Davis Food Safety Lab | California | University reference lab. For tremetol specifically — commercial labs may not have a validated method; UC Davis is the established reference. |
| U Penn Veterinary Toxicology Reference Lab | Pennsylvania | Alternative reference lab for tremetol and plant-toxin metabolites in animal products. |
Before sending any sample: call and ask whether the lab has a validated method for the exact compound you want tested. Not all labs can test for everything. A lab claiming a method but unable to produce method-validation documentation is a flag.
The split-sample strategy
Send portions of the same physical sample to 2 or 3 independent labs without telling any lab you are doing so. Convergent results across labs = high confidence in the analyte concentration. Divergent results = flag for follow-up.
This is the standard approach in any setting where a lab might, intentionally or otherwise, produce a result the operator wants rather than a result that reflects the sample. The cost is roughly tripled. The information gain is substantial.
Request raw chromatographic data
Ask the lab for the raw instrument output: LC-MS/MS chromatograms as .csv, .mzML, .mzXML, or the instrument's proprietary format. A legitimate ISO-accredited laboratory can produce this on request — the data lives on the instrument computer for a regulated retention period, and ISO 17025 requires it be available for audit. A lab that refuses to produce raw data is a flag.
Raw chromatographic data lets a knowledgeable third party verify the result independently of the lab's reporting choices: peaks can be re-identified against known compound signatures, reported concentrations can be checked against integrated peak areas, and contamination or interference peaks can be flagged. A computational analyst with no commercial relationship to the lab and no financial interest in the result is a useful adversarial check on any contested reading.
University-lab alternatives
Texas A&M Food Science Analytical Laboratory and UT analytical-chemistry facilities can perform food and water analytical chemistry with lower commercial conflict than national commercial chains. Results from academic labs carry distinct independent weight in any contested setting. They are typically slower and require more upfront engagement.