Strain Genie Helps Determine Which Product Chemistries are Right for You
Uniqueness is a virtue. Some people struggle for it. Others just are. When it comes to cannabis, our uniqueness can mean that what works for me might not work for you. And vice-versa. We’re individuals, and the explicit phytochemistry we ingest can interact with our unique physiologies in very different ways.
Some say that the nose knows when relating why we cherish specific scents. Our nose leads us to a specific scent, which is just a collection of different chemicals. When researching a book on terpenes, I discovered that many of the plants I love contain the same dominant terpenes. I favor those chemicals. Maybe, I need those chemicals?
Recently, I spoke with Nicco Reggente, Ph.D., of Strain Genie. Nicco hypothesized that there might be a connection between different cannabis chemicals and our genetic code. This idea of personalized cannabis medicine came to Nicco while he was finishing his Ph.D. at UCLA.
“I was simultaneously developing a machine learning skillset to predict patient treatment response and observing the proliferation of cannabis products available throughout Southern California,” Nicco explained. “At the same time, my wife was battling endometriosis and my mother breast cancer. I saw them both work up the courage to consider cannabis as a compliment to their treatment protocols.”
Their efforts were predominantly ineffective, however, which Nicco felt was understandable “given the near-infinite potential combination of cannabinoids and terpenes found in the thousands of cannabis products and cultivars, coupled with individual differences in one’s endocannabinoid system.”
As many cannabis consumers likely have experienced, trying to find products that work via a trial-and-error approach can be a futile endeavor, like trying to get a Honda Civic out of several inches of mud. The wheels spin and there’s smoke, but the mire is cruel, and the Civic succumbs. Nicco sought to incorporate his machine learning background such that better outcomes could be predicted, and the wheel-spinning and money wasted would be artifacts of times past.
“Leveraging similar algorithms to what I was using in graduate school, I helped build a recommendation system that took into account the medical conditions and symptoms an individual was trying to find relief for; their ratings of other products they’ve tried; and their similarity to other users that have tried additional products.”
Analogous to video streaming suggestions, Nicco’s algorithm could forecast which products should have high efficacy for an individual. Caveats to the algorithm’s functionality include if a consumer hasn’t tried enough products, and/or there are not enough individuals in the database that are like you with the same conditions.
“This is what’s known as the “cold start problem” in machine learning,” Nicco added. “Our foray into genetics was our way of combating this cold start problem. We were able to identify genetic biomarkers that could predict response to specific cannabis form factors (e.g. edibles, sublingual tinctures, inhalants) and identify cannabinoids and terpenes that would be most beneficial for conditions you may have a genetic predisposition for.”
Strain Genie uses two kinds of genetic biomarkers. The first set is directly related to “warnings” that can impact cannabis consumption and inform dosage, cannabinoid ratios, and form-factor recommendations. Examples of these biomarkers include lowering delta-9-tetrahydrocannabinol content in those with a genetic predisposition for cannabis-induced-psychosis; choosing form-factors that bypass first-pass-metabolism in cases of slow THC metabolism; or suggesting mindful consumption when there’s a risk of cannabis-dependence.
For example, knowing that you’re a slow metabolizer of THC can lead you away from edibles, which could make you feel much higher for much longer and possibly experience lethargy for several days after, when compared to other types of products. Furthermore, information that assists one in understanding the root cause of symptoms they experience is highly valuable. “Treating headaches that stem from eye pressure (e.g. from glaucoma) would require a different approach than headaches originating from Crohn’s disease,” Nicco added.
The second set is related to genetic predispositions for medical traits spanning 17 medical categories (e.g. cognitive health, sleep disorders, gut health) to assist in identifying root causes of the symptoms an individual may experience.
“Our service provides a bridge between peer-reviewed literature that has identified the genetic risk factors for specific conditions with research that has shown specific cannabinoids and/or terpenes to have been effective, at the very least, in ameliorating the symptoms associated with those conditions,” Nicco explained. “Our reporting process aggregates these insights, spanning 150 biomarkers and weights, each as a function of the severity of the risk, impact factor of the journal in which the findings were published, and the number of citations it has received.”
The end-result of the report is a personalized “cannabis cocktail”. Additionally, Strain Genie provides personalized recommendations within each of the 17 medical categories.
The trouble with suggesting cannabis products by their chic marketing names is that products can be genetically different than expected, rebranded for increased sales, or just improperly labeled. When constructing their algorithms, Strain Genie wanted to “mitigate the natural deviance of cultivars across batches, despite consistent naming.”
“Strain Genie’s recommendation is granular enough to provide insights as to the actual combination of cannabinoids and terpenes an individual should look for,” Nicco responded. This knowledge arms consumers to make more informed purchasing decisions. And by reading lab-test results and ingredient lists, consumers can identify which available products have the highest similarity to their ideal cannabis cocktail.
Strain Genie also relies on lab-testing data averaged across dozens of batches and discards those outlying results that deviate too much across batches from its product library. “Our product recommendation process (essentially identifying products with maximal similarity to the ideal cocktail) is the last step in a long process that produces all the necessary information for the consumer along the way,” Nicco added.
Some cannabis products may not agree with specific consumers. Many will, but users may embark on very different experiences. Some people report products that have made them irritable or euphoric or goofy or relaxed. Given that terpenes are cultivar differentiators [1-3], if a consumer knows which terpenes are needed for their unique body chemistry, they can choose a product wisely. Thus, Strain Genie also nominates terpenes that specific consumers should limit. “Terpenes in isolation and as part of the ensemble effect are huge differentiators in consumer-specific responses to products and are included in the cocktail that is recommended to users at the report-aggregate level and at the medical condition level,” Nicco explained. “Our summary report takes all medical categories into account in addition to medical condition-specific recommendations.”
So, back to that concept of the nose knows. I was keen to hear Nicco’s thoughts on an innate disposition for specific terpenes we require for distinct, genetic reasons. Can our nose target specific cultivars because we somehow know we need ingredients therein?
“We know from nutrition science that there exist several mechanisms that alter our food-seeking behavior and response to taste as a function of our gut’s microbiome,” Nicco answered. “We feel this through cravings (due to microbial influence on reward and satiety pathways) and altered taste perception (from microbial impacts on taste receptors) for certain foods, changing from day-to-day. As such, it is theoretically plausible that one’s nose could assist in the identification of the right chemotype for their current needs. Whether that is dependent on genetic factors is unclear, and further research would be needed in order to elucidate the ground-truth.”
Ultimately, knowing which products are correlated with our genetics can facilitate our medical usage of the plant for our individualized ailments. After all, many agree on the personal nature of cannabis, as the same product can elicit dramatically different reactions across individuals.
“Helping consumers understand themselves so they may better understand how to optimize their cannabis consumption is important for preventing negative experiences, identifying effective treatment protocols, and bolstering the public perception of cannabis as a viable supplement to a healthy lifestyle,” Nicco concluded.
- Orser, C. et al. “Terpenoid Chemoprofiles Distinguish Drug-type Cannabis sativa L. Cultivars in Nevada”, Nat Prod Chem Res, 2018, Volume 6: Page 304. [journal impact factor = 1.71; cited by 5 (Research Gate)]
- Fischedick, J. “Identification of Terpenoid Chemotypes Among High (-)-trans-Δ9-Tetrahydrocannabinol-Producing Cannabis sativa L. Cultivars”, Cannabis and Cannabinoid Research, 2017, Volume 2.1: Pages 34-47. [journal impact factor = N/A; cited by 12 (ResearchGate)]
- Reimann-Philipp, U. et al. “Cannabis Chemovar Nomenclature Misrepresents Chemical and Genetic Diversity; Survey of Variations in Chemical Profiles and Genetic Markers in Nevada Medical Cannabis Samples”, Cannabis and Cannabinoid Research, 2019, open access. [journal impact factor = N/A; cited by 2 (ResearchGate)]