AI is revolutionizing drug discovery by leveraging historical molecular data to identify new drugs and expedite treatments for challenging diseases. One pioneering company, VeriSIM Life, based in San Francisco, is spearheading this initiative. Their platform, BIOiSIM, features an extensive data repository of over 3 million compounds alongside 5,000 human and animal datasets, utilizing AI models to help pharmaceutical researchers virtually develop and test new compounds before initiating costly clinical trials.
In a significant development earlier this year, VeriSIM introduced AtlasGEN, a feature dedicated to biological translation simulations that assess a drug's effectiveness on human and animal bodies. This capability empowers researchers to prioritize drugs for further development and select appropriate animal models for testing.
“We can reduce animal testing,” stated Dr. Jo Varshney, CEO and founder of VeriSIM Life, during a recent interview. “We can refine experiments, testing drugs on fewer animals while validating results more computationally.” This approach not only enhances animal welfare but also optimizes research resources by preventing ineffective testing on incompatible species.
Dr. Varshney's passion for improving drug research stemmed from her early exposure to the pharmaceutical industry; her father worked in the field, and she pursued veterinary medicine before earning a PhD in genomics and computational sciences at the University of California, San Francisco. “I spent years exploring machine learning and asking how we can apply our understanding of medicine, biology, and chemistry to make innovative predictions,” she explained.
So far, VeriSIM Life has facilitated four clients in bringing their drugs to clinical trials, raising substantial funds from investors like Intel Capital and Village Global.
Challenges in Drug Research
The pharmaceutical industry, valued at approximately $1.6 trillion, faces significant hurdles, including skyrocketing R&D costs, which have increased tenfold in the U.S. since the 1980s, according to the Congressional Budget Office. Moreover, the failure rate for new drugs is alarmingly high, with around 90% failing clinical trials, despite an average development cost of over $870 million and a timeline of 10 to 12 years.
VeriSIM estimates that its BIOiSIM platform can reduce the time taken to transition from research and development to seeking FDA authorization for clinical trials by 2.5 years and claims to improve modeling accuracy by 82% compared to non-AI methods.
The Technology Behind BIOiSIM and AtlasGEN
VeriSIM Life’s BIOiSIM platform integrates sophisticated AI models and extensive datasets. “We employ machine learning methods, generative adversarial networks (GANs), and generative AI to identify new molecules within an immense space of 10^63 possibilities, narrowing it down to optimal molecular structures,” Dr. Varshney explained.
Crucially, VeriSIM has developed virtual analogs of various species, including humans, dogs, and rats, often utilized in clinical trials. They amalgamate knowledge from chemistry, physiology, and patient-based omics data to generate a score known as the Translational Index, akin to a credit score. This score ranges from 1 to 10, enabling researchers to gauge a drug’s viability and select suitable animal models for trials.
For instance, when testing a cholesterol-lowering drug, researchers can leverage the BIOiSIM AtlasGEN feature to analyze compounds and the Translational Index to determine optimal animal models, ensuring efficiency in their research efforts. “If it’s effective in animals but not in humans, the score diminishes,” Varshney noted.
The BIOiSIM and AtlasGEN platforms can simulate over 800 billion scenarios, tailoring their services to client-specific needs. Their team customizes animal models based on various factors, including potential drug toxicity in specific organs.
In terms of business strategy, VeriSIM Life takes a percentage of the revenue generated from drugs developed through its platform and also offers subscription-based pricing annually or on a project basis.
AI's Expanding Role in Healthcare
AI's impact on healthcare extends beyond drug discovery; innovative firms are employing generative AI models to assist doctors with cancer screenings, diagnoses, patient health predictions, and generating comprehensive applications for healthcare professionals and patients alike.
VeriSIM aims to significantly reduce costs and increase the success rate of drug trials with its BIOiSIM platform and Translational Index, ultimately improving and extending lives globally.