Hillel's Tech Corner: POMICELL predicts patient treatment compatibility

POMICELL utilizes AI to create virtual molecular models of real-life patients in order to simulate personalized drug reactions.

POMICELL utilizes AI to create virtual molecular models of real-life patients in order to simulate personalized drug reactions. (photo credit: Courtesy)
POMICELL utilizes AI to create virtual molecular models of real-life patients in order to simulate personalized drug reactions.
(photo credit: Courtesy)
Anyone who has ever experienced a serious skin condition, or many other types of medical conditions for that matter, is well aware of the phenomenon that the treatment you are given simply doesn’t work.
I’ve discussed this very issue in the world of mental health and how all treatments are basically trial and error. Now why is this a problem? Because while the psychiatrist or the physician prescribes a certain medication, you might be suffering for weeks only to find out that you did not respond to that specific treatment and you have to start the process again. There has to be a better way.
Imagine if we could create a virtual model of real people, with their unique backgrounds and conditions, and simulate their reaction to specific treatments in order to determine the best fitting treatment for them, thus optimizing the effectiveness of delivering care.
Meet POMICELL, a company founded in 2016 that utilizes AI to create virtual molecular models of real-life patients and various types of drugs and diseases, in order to simulate personalized drug reactions and accurately predict patient to treatment compatibility.
The company has 15 employees and eight experts working for them and is based in Jerusalem. As far as capital, POMICELL has operated on a budget of $3.5 million to date, received from angel investors, grants and company revenue.
POMICELL is unique in its ability to automatically combine scientific Big Data found online with patient data from personal electronic medical records to create accurate disease models of specific tissues, and tailor them to every patient. Moreover, unlike many AI-based technologies found in various industries, POMICELL offers elaborate and intuitive explanations of the results they achieve, allowing for better implementation by researchers, doctors, and regulators.
The company’s vision is an ambitious one: to enable any patient in the world to be subscribed to their virtual tissue data-bank, and be able to predict success or failure of a specific drug or substance, in order to get the optimal treatment for them.
POMICELL is able to take small amounts of medical data of each patient, and enrich it with Big Data by using their unique AI capabilities, in order to create a personalized tissue model per patient. These personalized models create virtual OMICs libraries per patient and enables full-scale clinical simulations of treatment outcome.
The company’s current business model is to assist pharma companies with optimal selection of real patients for the clinical trial (they are already commercial with pharma companies). Success rates of clinical trials are less than 10% (to get from clinical trial to actual marketization) and so the risk is great for pharma companies to get new treatments to the market. The main reason for failure is the inability to predict the clinical trial participants’ response to the treatments.
POMICELL ALLOWS pharma companies to do virtual trials on actual patients. POMICELL has modeled more than 100 types of diseases. Over the last two years, POMICELL provided services to pharma companies mainly in the field of dermatology, stem-cells and cancer, which are the company’s current focus.
As far as the team, Boaz Buchandler who is the founder and CEO of POMICELL, is an expert in bioinformatics, DBA, data mining, and customer relationship management. Prior to POMICELL, he spent seven years leading R&D processes, developing software solutions, and managing products in early-stage data-oriented start-ups, mainly in the biotech industry. Buchandler is an eternal volunteer and a member of an initiative seeking solutions for rare diseases using bioinformatics. He holds a BSc in biochemistry and an engineering degree in computer programming from the Technion.
Ouri Fischel, the company’s COO, brings to POMICELL 15 years industry and leadership experience in biotech and drug development companies. Specializing in bioinformatics, cheminformatics, in-licensing technologies, as well as strategic planning and partnerships,  Fischel’s industry track record includes collaborations with Sanofi-Aventis and J&J, and business development at Synergix.  Fischel holds an MSc in Biology from The Hebrew University of Jerusalem and an MBA with a specialty in biomedical management from The College of Management Academic Studies.
Finally, Amram Ben-David, the company’s CTO has 15 years’ experience in Full-Stack development and database analysis. Ben-David led programming teams in well-known hi-tech companies and start-ups specializing in cyber, fintech and big data applications. Also, he served as an officer in the IDF’s Intelligence and Information Systems.
It is important to mention that POMICELL, like many companies, is also using its technology to better the treatment of COVID-19. Their technology is using its artificial intelligence solution to create a personalized novel coronavirus (COVID-19) infected epithelial cell tissue computational model that aims to enhance the effectiveness that the new therapies/vaccine a for COVID-19, as well as treatment combinations, may have on population, shedding additional light on important mechanisms of actions of the COVID-19 cells to treatments.
Their ability to build models that are both deep and personalized, automatically allows them to make ongoing improvements to the models based on feedback from the field and match patient groups with the optimal treatment in accordance with their condition.
If you think about what POMICELL is doing it, in essence, really is borderline science fiction. The ability to virtually simulate the human tissue of a person in order to determine whether a certain treatment will work is some pretty mind-blowing technology and I, for one, look forward to this technology being widely adopted and accessible.