1. Help Center
  2. Microbial Safety Testing

Did the Michigan Marijuana Regulatory Authority (MRA) ban molecular methods for microbial testing?

No, the MRA did not ban qPCR. In this article we aim to clarify some items highlighted by the MRA in their recent memo on qPCR.

Did the Michigan Marijuana Regulatory Authority (MRA) ban qPCR?

Does this mean my Total Yeast Mold/Total Coliform results up until now have been wrong?

  • No. The MRA’s memo and new requirements mean that labs must provide additional validation data for TYM and TC methods if using qPCR for total count tests. It says nothing about past results.
  • This announcement does not affect qPCR testing for presence/absence tests.

Have I been releasing dirty samples into the market?

  • Absolutely not. qPCR is superior to plating in detecting species of yeasts and molds that are dangerous when consumed or inhaled. Please note that some states do not require any total count testing at all, and rely only on species specific microbial testing using molecular methods.

Are your Cq/CFU enumeration equations accurate?

  • Our extensive validation efforts show our development work on our Cq to CFU conversion equation. This has always been public and is also published in two peer reviewed papers and one non-reviewed white paper.
  • The issue that we believe precipitated the MRA's actions started with a new NSI CRM for TYM proficiency testing.  This PT was developed for culture based methods and does not produce qPCR results that are consistent with plating using MGC's SOP for total count tests.
    • Our own investigation found that the NSI CRM for TYM (the emerald test panel) was  ground to a fine powder, frozen, and contained cryopreservatives that are known to inhibit gDNA extraction.
  • In general, this is not surprising since plating for “total” counts are inherently flawed and not comparable to qPCR results. Not all yeasts and molds grow in culture plates, but are picked up in qPCR. Bacteria often grow in TYM  plates, but our assay is designed to exclude the bacteria. We write about this extensively in our help center:

Is the TYM assay accurate? 

  • Yes. Our method validation for total yeast and mold is more robust than any on the market and the data is accurate. The Cq/CFU equation in our validation document was generated with samples spiked with live organisms. Those organisms were chosen because they are well documented to be found on cannabis, or were well studied yeasts/molds,  and in most cases were used in 3Ms own method validation.
  • Our most up to date inclusivity and exclusivity document has always been public. A current list with updates from our current additional validation efforts can be found in our help center here:

What is MGC doing to address the situation with total count tests in Michigan?

  • We are already well underway with a full method validation ourselves following AOAC Appendix J guidance as stated by the MRA. We will publish our updated validation of our method once complete and provide notice to our customers. 
  • We are a full participant in an AOAC Emergency Response Validation that will hopefully soon result in AOAC certified methods (culture and molecular) for the testing of TYM on cannabis.
  • Once complete our new standalone TYM manufacturer method validation update and the inclusion/exclusion criteria provided within should suffice for any safety testing lab in Michigan to show the MRA they have sufficient evidence of a compliant method. 
  • We again encourage you to contact the MRA directly about what they will require.

Are there any changes coming to the TYM or TC assay?

  • No. We are not making any changes to the primer design for our TYM or TC qPCR assays. 
  • We are investigating what changes we can potentially make to our extraction method so that we lyse open more cells, improve genomic DNA recovery, and improve concordance with plating when compared with industry CRMs.
  • These changes may result in a protocol update for the extraction step  for the TYM assay, but no change to the assay design itself.