Cellworks: Ideal Treatment Selection with a Computer Predictor

Shireen Vali, Co-founder & CSO and Taher Abbasi, Co-founder & COO

It started with a simple question. Why doesn’t the development of new drugs look more like the design of new silicon chips? After all, no electronics company would risk the expense involved in building a new chip without first validating the design through simulation. So why does the pharmaceutical industry still depend on the trial and error method—conducting trials on patients to find out how well a drug works?

This was the conversation between Taher Abbasi and Dr. Shireen Vali, a veteran electronics engineer and a leading biologist, and the husband and wife team behind Cellworks. Thirteen years ago, they began a pioneering effort to predict the efficacy of drugs through computer simulation. Given the bewildering complexity of human cells, building computer models to simulate the myriad chemical pathways and their interaction with drugs is a herculean task. Adding to the challenge, many diseases, including cancers, are the result of multiple genomic aberrations that vary from individual to individual, meaning that any meaningful simulation must analyze the efficacy of a drug for a specific patient.
With the backing of legendary Silicon Valley investors Sequoia Capital and Artiman Ventures, Cellworks built a multi-disciplinary team of molecular biologists, mathematicians, and software engineers in Bangalore to develop a set of patient and disease models that can predict a patient’s response to different therapies.

The result is products now being evaluated by oncologists around the globe that provide a Therapy Response Index (TRI), a ranking of possible treatments analyzed through AI-driven biosimulation, that helps doctors select the most efficacious therapy for their cancer patients. “Essentially, Cellworks uses TRI to perform the trial and error processing, simulating in software the response of an individual patient to an array of potential therapies,” says Dr. Vali, “We want to help the physician to select the best therapy first, to save the patient precious time, and to avoid the debilitating effects and cost of rounds of unsuccessful treatment.”

Early identification of the most efficacious treatment is a significant advance for Precision Medicine. Today, doctors have access to detailed genetic testing but relatively little guidance on how to select therapies. The Standard of Care identifies several approved therapies, but the physician does not know which therapy is best for which patient. Cellworks’ Singula product uses the patient’s molecular fingerprint to predict the therapy that will provide the best outcome.
In cases where the patient does not respond to any Standard of Care option, Cellworks’ Ventura product opens the TRI lens to consider all FDA-approved drugs and can identify combination therapies of multiple drugs working together to produce a favorable outcome.

Cellworks is answering that original question about drug development, too. “We now have the technology to identify patients who are likely to respond well to investigational drugs before they have been tested in humans,” says Taher Abbasi. “We know that most cancer therapies will only work for certain patients, and we can apply modeling capabilities to predict which receptors to target, which indications to target, what percentage of the population will respond, select likely responders for clinical trials, and rescue or repurpose drugs.”

Abbasi highlights the fact that Cellworks solutions have been tested in clinical studies where the computer predictions have been compared to the patient outcome with excellent results. The results provide hope for improvement across the healthcare ecosystem: Better outcomes for physicians and their patients, reduced cost to payers and healthcare providers, and lower drug development costs for pharmaceutical companies. Engaging with payers in the U.S. and the UK market, the company is expanding to China and India. “We’re expanding into new cancer indications and drug classes to address clinical unmet needs,” concludes Abbasi.