In what BenevolentAI considers a move that deepens its drug discovery and development expertise on its board, Dr John Orloff has been appointed as a Non-Executive Director. Dr Orloff joins the BenevolentAI board as it advances plans to scale its AI platform, partnerships and drug portfolio. DDW’s Megan Thomas speaks with Dr Orloff about his career and the future of AI in the industry.
MT: What is your background in drug discovery and AI?
JO: My background expands across research, drug discovery, clinical development, regulatory affairs and health economics. My most recent operational role was as EVP and Global Head of R&D at Alexion, and I previously held executive R&D roles at Baxalta, Novelion, Merck Serono and Novartis.
In recent years, I’ve seen new technological advances dramatically change the way the industry approaches drug discovery research. It’s inspiring – and exciting – to think about how this new category of companies are building tools and technologies that can help deliver medicines to patients faster, more accurately and at scale. It’s essential to bring scientists and business leaders with extensive drug discovery experience to the table to help shape this future of drug discovery, which is where I fit in.
MT: Can you provide an overview of what BenevolentAI does, and what your role will be at the company?
JO: BenevolentAI is a leader in the field of AI-enabled drug discovery. Benevolent builds “tech in the service of science”, which means developing computational tools for scientists, with scientists. Benevolent integrates AI across the R&D cycle – from target identification to patient stratification and chemistry – to enable scientists to discover the best therapeutic intervention. Its approach allows scientists to make higher confidence decisions – a critical step in helping to eliminate the well-published and costly mistakes that have traditionally plagued pharma R&D downstream. The company has developed an impressive AI-derived in-house pipeline, and my role as Non-Executive Director will be to serve as a sounding board as Benevolent deepens its R&D capabilities and advances these programmes into the clinic.
MT: BenevolentAI has demonstrated in silico, in vitro and clinical validation of its technology. How does this approach differ from the drug discovery and development industry as it currently functions?
JO: Researchers have traditionally viewed data in silos, and even within AI-applied drug discovery efforts, companies are still focused on integrating disease-specific data sets, which is too simple to truly represent the complexity of disease biology. BenevolentAI is different – it integrates all available data from a range of modalities and therapeutic areas to give scientists a holistic view of all the information that’s available. With this data-engine in place, the Benevolent Platform starts with the most critical stage in the process: finding the right target. The Platform then enables scientists to refine hypotheses, validate targets and synthesize and test molecules. Ultimately, this approach enables high-confidence drug candidates – optimized by target, mechanisms of action, and patient subtypes – to progress through to the clinic.
MT: What excites you most about what BenevolentAI is currently achieving, and where do you think it has room to grow?
JO: What’s really interesting about BenevolentAI’s approach is the idea of being able to be ‘right more often’ and remove some of the empiricism that is intrinsic to traditional approaches to drug discovery. When you look at Benevolent’s approach to data, and combine that with its target-ID and precision medicine capabilities, you can see how scientists are really able to define and refine the right hypothesis from the outset. This hypothesis underpins the success of the entire drug discovery process, so it’s a radically stronger approach to traditional drug discovery. Benevolent has an advanced clinical pipeline, with candidates in pre-clinical and clinical development, and it also has a number of Early Discovery Programmes. It’s really exciting to think about how the increasing sophistication of Benevolent’s tech will impact the development and rapid progression of these 20+ early drug programmes. And down the line, many of the targets already identified can be leveraged as platforms to branch out to new disease areas that can exponentially expand the impact of this technology.
MT: Has the pandemic changed the need for AI in drug discovery and development? To what extent did AI play a part in the development of vaccines and therapeutics for the virus?
JO: The pandemic has certainly brought the challenges of developing effective therapeutics into sharp focus. Almost two years in, as vaccines are being rolled out across the world, the search for safe and effective treatments continues. Yet, physicians do have some effective candidates in their armamentarium. One such drug is baricitinib – first discovered as a treatment by BenevolentAI using its computational tools and AI-enhanced knowledge graph. This technology identified baricitinib as having anti-cytokine and anti-viral properties. Baricitinib is now approved for emergency use in the US, Japan, and India after being validated in global clinical trials, where it was shown to reduce deaths by 38%. This discovery, which was conducted over a single weekend, demonstrates that AI will play a significant role in the pandemic response, be it in accelerating the search for new treatments, or in pinpointing, with greater accuracy, the treatments that could be most effective in preventing mortality. More broadly, AI will help to accelerate, with greater accuracy, the discovery of new treatments for a wide range of devastating and life-threatening diseases, for which there remains significant unmet need.
MT: What are the three main benefits of AI in drug discovery, what are the challenges to uptake and how do you feel they can be overcome?
JO: AI can help to deliver medicines to patients faster, more accurately and at scale, but companies must be accountable for the way they communicate their advances. Large pharma companies recognise the benefits of AI, but over-hyped claims have undermined trust and slowed adoption. The key is to adopt a realistic, proof-point led approach to communicating progress. For example, a company like Benevolent is certainly not short of milestones, so it’s all about showing the world the tangible benefits of integrating AI and how these approaches can, ultimately, help improve patient outcomes.
