Computer Analysis and Data Mining Bring Scientific Basis to Medical Marijuana

Fractal Realms series. Backdrop of fractal elements, grids and symbols on the subject of education, science and technology

Computer Analysis and Data Mining Bring Scientific Basis to Medical Marijuana

Laboratory analytics coupled with data collection and data mining are yielding valid information on the clinical effects of cannabis and its wealth of associated molecular materials. The results are starting to appear in the form of products that will be available to patients, doctors and growers.

by Tom Williams, Editor-in-Chief RTC Magazine

As the acceptance of medical marijuana grows with 26 states now having legalized it and more to come, the awareness of the health benefits of marijuana’s various compounds is growing and maturing as well. The field is still struggling to find a systematic way to characterize the differences between these compounds, called cannabinoids, contained in the various strains of the plant in order to accurately apply them to the various symptoms for which they have generally been found to be effective. In addition, of course, medical research continues in countries where it is currently allowed, namely Israel, the Netherlands and Brazil. Still there is much work to be done to optimize the already recognized medical advantages of cannabis and to characterize and apply even more. While there are some 30 molecules, known as cannabinoids, present in cannabis, only one, tetrahydrocannabinol (THC), is known to have psychotropic effects. Many of the others are proving to be effective against a host of ailments including anxiety, schizophrenia, HIV, arthritis, and possibly even cancer.

One company that is applying electronic diagnostics and data mining to the task is PotBotics, based in New York and Palo Alto, California. According to its CEO, David Goldstein, the company is seeking to educate the market by better understanding and quantifying the effects of cannabis both from a cultivation point of view and the actual neurological point of view. In fact, the company is working on three tools to address three major points of medical cannabis: the dispensary/patient relationship, doctors and growers.

Currently, local dispensaries work in individual agreements with laboratories such as SCLabs in Santa Cruz, California, to obtain the cannabinoid profiles for the strains they offer. SC Labs tests for eight different cannabinoids and is gearing up to produce identification and quantification further elements in these strains as well. With knowledge of the cannabinoid profiles, it is then the job of the person known as a “budtender” to determine which strains are best suited for a given ailment or combination of ailments. A budtender acts as an advisor, as a kind of sommelier, who draws on knowledge gained with extensive experience and interaction with patients and the different strains and their reactions to them. Most physicians, even those who are convinced of the medical benefits of marijuana, do not have the depth of knowledge to be able to select specific strains for specific treatments. They rely on the dispensary’s budtenders to supply that.

With the information on the effects of different cannabinoids and ratios of different cannabinoids growing, the ability of individual budtenders to master the volume of data has its limits. According to Goldstein, there are over 750 strain names and most of these are marketing/branding names that frankly sound silly in a professional context. No MD is going to seriously want to prescribe “Maui Wowee” for a patient. The patient today is left to take the prescription to the budtender and rely on his or her still considerable knowledge to receive the strain that will be most effective for the particular condition. In addition, if a doctor were to specify a specific strain, it might be one that is not available in that area. So the more attractive way to address this is to be able to search for the available strain that best meets the requirements. PotBotics is approaching this with its first product, called PotBot, which Goldstein describes as a “virtual budtender.”

PotBot will be available as an online service for desktops, tablets and smartphones, but will also be offered as a kiosk for use by patients within dispensaries (Figure 1). The patient, who in most states must have a doctor’s prescription and a medical marijuana card issued by the state, enters age, gender and weight and then selects an ailment and the system displays the optimum combination and proportions of the recommended cannabinoids. These are based on the analysis of cannabinoid levels coupled with the latest scientific data, which is constantly being refined by PotBot’s Cloud repository and data mining operations.

RTC12 ERTW PotBot Fig1

Figure 1: The PotBot virtual bud tender as a kiosk that will be placed in medical marijuana dispensaries.


Since there is a chance that locally available strains will not contain the ideal combinations of the molecules recommended, the system then displays one or more available strains that most closely match the recommended list. It lets the patient examine the exact ratios of cannabinoids in each strain to make an informed choice (Figure 2). If the system is a kiosk in a dispensary, the selection will naturally be from strains available at that location. If it is a desktop or mobile app, the patient will be presented with a list of dispensaries in his or her area with maps and directions.
Doctors who do understand the medicinal value of cannabis are frustrated by the lack of research and scientific data for treating specific symptoms. The information presented is based on PotBotics’ “cannabinoid matrixes” or the specific ratios of the various chemicals put together from analysis, research and PotBotics’ own Cloud-based data collection and artificial intelligence-based data mining. The correlation between cannabinoid ratios and their appropriate target symptoms is constantly being refined. One major tool in this activity is an FDA-approved electroencephalogram (EEG) helmet called BrainBot that can be used by doctors to measure patients’ responses and supply data for further evaluation. See more about BrainBot below.


RTC12 ERTW PotBot Fig2

Figure 2: A screen in PotBot showing a selected cannabis strain and the concentrations of six of the major cannabinoid compounds found in that strain. Note the low level of the psychotropic THC.

The current version of PotBot only allows the user to enter one ailment or symptom. According to Courtney Codd, marketing coordinator for the dispensary The Cookie Co. 831 in Soquel, California, a large number of their patients are looking for strains that can address combinations of ailments. Goldtstein acknowledges this need and has stated that although the current beta version of PotBot does not support multiple symptoms, that capability is currently under active development.

As noted above, further research is examining other compounds in the plants that can influence the effects of the cannabinoids. This has led to the growing recognition of the importance of terpenes, which was emphasized both by Codd and the dispensary’s budtender, Joe Doruff. Terpenes, many of which are aromatic hydrocarbons present in most plants, influence the smell and flavor of cannabis and are now being recognized for their ability to modulate and modify the effects of cannabis. The current beta of PotBot does not now include terpenes, says Goldstein, because there is not yet sufficient data that can be correlated with their effects to make a useful tool, but that inclusion of terpenes is definitely on the roadmap.

Data from and for Doctors

As noted earlier, even those doctors supportive of medical marijuana are frustrated by the lack of research and scientific data on its use and effects on patients. Potbotics’ BrainBot is an FDA-approved EEG helmet that lets physicians capture brain waves with the protocols used by pharmaceutical companies to evaluate their offerings—effects on blood pressure, epilepsy, anxiety and others. That same EEG data can then be uploaded to the Cloud to be combined with data from other patients and then subjected to data mining to see the matrix of effects on that and other individuals compared with a list of strains and consumables to establish correlation between those strains and the most appropriate symptoms to which they can be applied. That same data can be made available to other researchers. Again, the collection of such EEG data is growing and may not be immediately generally applicable to treatment but as the amount of data and the matrix of correlations grows, more accurate and effective treatments will become possible. The use of BrainBot is also intended to yield information on the effect of dosage levels. So administration of cannabis is not done through smoking but rather by means of a tincture and an inhaler, which make possible more exact dosage and also yields faster effects.

Data for Cultivation

The next phase for PotBotics will address the third aspect of the industry: breeding and cultivation. Although not yet available, the third product will be called NanoBot, an advanced DNA reader that will look at the genetic fingerprint of actual cannabis seeds. NanoBot will have two major objectives, both of which will develop from the DNA analysis correlated with the data being collected on cannabinoid/terpene content and patient data. The output for the grower will on the one hand obtaining a specific growth plan for a given strain based on the optimal amounts of irrigation, sunlight, fertilizers, etc., for growing specific strains and hybrids. The DNA data will also help guide growers toward breeding strains with specifically targeted cannabinoid contents aimed at specific treatments.

Like the history of medicine itself, the development of cannabis for effective medical treatment is passing from a stage of lore and anecdotal knowledge—much of it quite effective—to a stage based on accurate data and research. As the volume of data grows, so will the more accurate and effective application of the beneficial clinical effects.

PotBotics, New York, NY. (917) 525-2697.

SC Labs, Santa Cruz, CA. (866) 435-0709.

The Cookie Co. 831, Soquel, CA. (831) 471-8289.