We got to talk with Kevin Mckernan from Medicinal Genomics, a company that tests the DNA of cannabis strains. |
While he may not be a household name, Kevin Mckernan's work sequencing the genomes of cannabis strains is incredibly fascinating. Kevin used to work on the Human Genome project at MIT and later applied his work to cannabis. He went on to co-found Medicinal Genomics and Kannapedia, which have given significant insight to the cannabis plant's family tree. Medicinal Genomics' latest addition is StrainSEEK, a service to identify and register strains, getting growers prepared for the patenting issues that arise as more states pursue legalization.
CannabisTech: You’ve had a very interesting career so far. How did you go from being the R&D lead on the Human Genome Project to starting the Cannabis Genome project?
Kevin Mckernan: As the human genome came to close, we decided to spin many of the technologies we developed to sequence the human genome into a company called Agencourt Bioscience Corporation.
This company developed DNA purification nanoparticles and robotics while also becoming the largest commercial sequencing center from 2000-2005. In 2005 Beckman Coulter Acquired the business but didn’t know how to value a skunks work project we had to change the face of sequencing. We spun this 19 person company out in the course of this acquisition as Agencourt Personal Genomics. This small team developed the SOLiD sequencer which was acquired by Applied Biosystems in 2006. Applied Biosystems (ABI) was ironically the company that funded the private Human Genome Project at Celera Genomics and was the market leader in DNA sequencing for most of the industries existence.
The SOLiD Sequencer was 100,000 times faster than the ABI sequencers and I worked for ABI from 2006-2011 to push the technology into the marketplace. After we captured 35% of the market Life Technologies acquired ABI in 2008 and we began to use these sequencers to sequence patients tumors. If you sequence a tumor and the patient's genome you can find smoking gun mutations that are candidates for more targeted therapies.
This method captured the cover of Science Translational Medicine and I soon began to get inquiries from friends with cancer asking about cannabinoids. (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2858564/pdf/nihms-192215.pdf)
I was only aware of palliative effects of cannabinoids at the time and after digging deeper into Guzman and McAllister's work I realized we had safest anti-tumor compound handcuffed by reefer madness. Much of the reefer madness in the medical field was referring to a feature, not a bug. The incredibility diversity in chemotypes between plants was held up as a problem when it's, in fact, a medical opportunity and with more measurement on the plant, we could put that straw man argument to bed.
So we consulted a few attorneys, spent a few weeks designing a DNA purification lab I could fit in a suit case and get through the TSA and flew to the Dylon Hotel in Amsterdam and purified 10 strains and brought the DNA back for sequencing.
Kevin working from the lab he packed in a suitcase at the Dylon Hotel in Amsterdam.
CT: Verifying the legitimacy of strains is more important than ever now that cannabis strains can be patented as intellectual property. Have you come across any court cases or horror stories involving growers that could have been prevented by a service like StrainSEEK?
KM: The initial patents that have issued are both very broad but also easy to navigate. I'm not an IP attorney so always confirm what I say with an expert but in my experience over 21 years of patent litigation, there is an asymmetric cost against the defendant in IP cases. The cost to invalidate a patent is over $100K through the ex-parte re-exam process. The USPTO makes just as much money on the invalidation process as they do on the patent issuance process so there is not a strong incentive at the USPTO to get the prior art process perfect. Hence many broad patents issue where there is clear prior art and the burden falls on the defendant to invalidate. This invalidation cost can be short circuited with Prior Use Exemptions in the US plant patent law. If you are using a strain 1 year in advance (in a business setting), you have an exemption to any submarine patents that emerge in that time frame.
Given the federal illegality of Cannabis, I think most companies that have federal cannabis patents are sitting on them until legalization occurs. Nevertheless, the issuance of these patents will chill investors looking at companies using genetics that openly patented. At the moment it appears the Backes patent covers all type II plants. Greater than 3%THC and 3%CBD is covered. I am skeptical this was novel in 2013 but until someone proves otherwise, they own that class of plants.
GW navigates this by growing Type Is and Type IIIs in different grow rooms and mixing the extracts but for those baby boomer pain patients that are attracted to type II flowers, there is soon to be a legal fight over this.
Had Type II plants been strainSEEK’d in 2013, this would have created a prior art and limited the scope of that patent.
CT: If I understand the process correctly, StrainSeek allows growers to send samples to Medicinal Genomics’ partner labs around the country, where their genome is mapped and uploaded to Kannipedia, where growers can see their results. How long does it normally take to map a sample’s genome and upload the results? (Let’s say for this scenario, that the grower just dropped their sample off at the lab so we don’t have to worry about accounting for shipping time)
KM: Sequencing is 2-4 weeks once the DNA is in house. We now have stem kits that allow direct shipment to Medicinal Genomics. We run our own Sequencers so we are providing 25X more data than any other provider in the space. 3.2 Million bases are sequenced. We have a credit program that encourages growers to work with a local lab to get terpene profiles shipped in.
CT: How long would something like this have taken ten years ago? (If this was even possible ten years ago).
KM: It was possible 10 years ago but the costs were exponentially more expensive. The SOLiD sequencer came to compete with the ILMN sequencer in 2007.
The cost of genome sequencing has gone down significantly in the last ten years.
CT: Was there a particular technological advancement or combination of advancements that made StrainSEEK possible?
1)More competitive sequencing costs. We played a roll in developing this with both the SOLiD sequencer and the Ion Torrent Sequencer.
2)We sequenced the first Cannabis Genome and put it public on August 18, 2011
3)Field compatible DNA preps so people can ship DNA in the mail. SenSATIVAx allows this.
4)We also developed a Stem kit that allows legal shipment of stems according to DEA vs Hemp Industry Association.
5)Decentralized Databases like Bitcoin. Relying on Federal databases for your cannabis DNA is unwise. These get shut down every time there is a budget crisis. (See this study: The 2013 US Government Shutdown and Health: An Emerging Role for Social Media)
CT: Observing DNA through a powerful microscope, then using a machine to sequence it, and uploading it to a database almost sounds like science fiction. I’ve always been especially curious about how the information in DNA is transferred to a digital format. Without giving away any trade secrets, could you walk me through this process? I suppose this would be “Sequencing DNA and uploading it to a database for Dummies”
KM: DNA Sequence information is really derived from a Signal on a CCD camera like your phone. A group of pixels lights up with 1 of 4 colors and we can assign those colors to ATCG (Red=A, Yellow=T, Green = C, Blue= G).
Once we have ATCGs those files are compressed and uploaded to various databases and looks something like the following:
This is a lot of data so many people like to compress the data using a Variant Call Format (VCF) which only documents the variants in the DNA based on the Reference genome used. This is basically a list of your strains uniqueness and hence the VCF files are often used for IP and for Hashing into the Bitcoin Blockchain. The Blockchain hash is the most ironclad proof of existence. No one can reverse this transaction and it's digitally etched into over 10,000 databases all over the world. If Kannapedia goes down, you still have your VCF file and your Hash in the blockchain to prove your file was generated at this given time.
CT: Most companies in the cannabis industry are headquartered out West, but Medicinal Genomics’ headquarters are in Wodburn, Massachusetts. Are there any advantages to being located there?
KM: Boston is one of the few Genomics, Medicine, and Biotech capitals of the world. Massachusetts is the Colorado of the east coast in terms of legalization. Not quite as mature as CO but we are the most likely state to get there first in New England. We are the largest legal market, close to Toronto. A flight to Amsterdam and Spain is equidistant to a flight to CA. Cannabis is much more international today than just the west coast.
CT: Lastly, what comes next for StrainSEEK? Are there any updates or new additions to StrainSEEK that you’re looking to add in the near future?
KM: StrainSEEK is getting a 5X upgrade this fall. We now Sequence 3.2Million bases (used to be 850,000). Other providers on the market only sequence 120,000 bases. If you want to find uniqueness, more sequence is better than less. This is important for IP. We also sequence over 2 dozen genes in the cannabinoid and terpene synthase genes. Most other players don’t cover any. The third upgrade is our Rosetta stone panel. Built into StrainSEEK Version 2 is targeted DNA that has been sequenced by others in the field. We cover 850 SNPs from our Version 1 data. 300 SNPs from the Lynch/Kane study. 300 SNPs from the public Phylos dataset and 50 SNPs from the Sawler paper. We can uniquely tell if your strain is unique amongst all things that are public as of 2017 and no one else can do that today. This is very important for IP as there is so much public data now but until now, everyone was sequencing their own pet regions of the genome. None of the data overlapped for cross comparisons and StrainSEEK V2 puts an end to that.