AI-Driven Predictive Analytics: Transforming Drug Development

Conceptual technology illustration of artificial intelligence  

(Adobe Stock 296043415 by kras99)

Many health care firms are integrating artificial intelligence (AI) into their operations and services. A specific healthcare technology company, for instance, has created a biosimulation platform driven by AI. This platform empowers the pharmaceutical industry by accelerating and improving outcomes in drug development, thereby contributing to the progress of human health. To learn more, Infection Control Today^® (ICT^®) interviewed Jo Varshney, PhD, founder and CEO of VeriSIM Life.

ICT: Can you provide an overview of how your health care company implements AI technology in its operations and services?

Jo Varshney, PhD: VeriSIM Life is a technology company that delivers predictive analytics services to our clients using AI that forecasts the probability of a drug candidate’s clinical success—the insights provided help accelerate drug development and reduce unnecessary experimentation. The company’s computational platform is built on machine learning and quantitative methods to bridge therapeutic translation from lab to clinic.

ICT: Could you share examples of AI applications or solutions that have been particularly successful in your health care company?

JV: Our clients’ success drives our success. While each project we deliver is unique to a client’s requirements, we’ve helped shave years off preclinical development and millions in unnecessary testing. In one recent engagement, our client asked us to help sort through dozens of combinatorial oncology treatments to determine the most efficacious cocktails. A matrix of dosing regimens was created for each combinatorial therapy to evaluate the predicted combinatorial efficacy at different dosing regimens. A combinatorial therapy’s dosing matrix was analyzed to establish key metrics, eg maximum and average inhibition predicted across a dosing matrix. This resulting analysis—completed in just 2 months—ranked the combinations regarding efficacy, allowing our client to focus on only the optimal candidates and saving many more months in expensive in vivo testing.

ICT: What challenges or obstacles have you encountered in integrating AI into health care, and how have you addressed them?

JV: The use of AI in drug development is still in its early days, and regulatory science related to its integration is still developing. Agencies like the FDA are working with companies like VeriSIM Life to publish guidelines and standards. Working in a formative period such as this can be challenging. Still, we have focused on proactive communication and collaboration with the FDA, engaging directly with the agency across multiple channels and with other entities such as NIH, EPA, and National Academies of Science.

ICT: How do you ensure that AI systems in health care adhere to the highest patient data privacy and security standards?

JV: There are continued concerns with data quality, security, and algorithm bias. Still, at VeriSIM Life, we’re addressing these by establishing several benchmarks to confirm our continuous system validation simulations and designing explainability into our systems. This allows us to reduce bias and ensure that results are reliable. Ultimately, using AI does not replace good science or the rigorous process of ensuring drug safety. We believe it should complement these processes.

ICT: In what ways has AI helped your organization streamline administrative processes and reduce operational costs?

JV: Our staff has deployed AI in several operational workflows, including using it to assist with the generation of reports and to inspect datasets that our platform ingests to prevent errors, inconsistencies, and unintended biases. This has allowed our team to be more strategic in supporting client projects and has reduced the overall operational burden.

ICT: Lastly, what advice or key takeaways would you offer other health care organizations looking to leverage AI effectively in their operations and services?

JV: My biggest advice would be to invest in AI today. Now is the right time to embed AI into your drug development processes to make them more cost- and time-efficient. I appreciate that organizations that are not “AI-native” can be reticent to adopt AI techniques, and employees are often resistant to leverage it, perceiving their jobs could be threatened by AI replacement. But leveraging AI is about adapting to change, which we’ve embraced in business for decades. So, I would encourage organizations to develop competency internally and seek partners with a pedigree of integrating this new technology into their clients’ workflows instead of replacing those workflows wholesale.