Artificial intelligence and machine learning are beginning to revolutionise healthcare.
AI-enhanced review of medical imaging has proven adept at flagging things the best-trained eye might miss.
Natural language processing is helping alleviate the burden of EHR/EMR input on healthcare professionals.
Predictive analytics are being used to identify individuals at risk of developing chronic diseases, allowing for early intervention, or to identify the treatment options with the greatest likelihood of success.
And we are now beginning to see the first AI-generated drug candidates enter clinical trials.
These technologies promise a transformation in how pharmas and biotechs develop and commercialise innovative treatments.
It’s early days yet, and there are formidable challenges – regulatory and technological alike – to be overcome.
However, despite the obstacles, the life sciences industries are already starting to realise the potential.
Simplifying search
Since OpenAI’s ChatGPT debuted in March, there’s been a lot of discussion of the tremendous potential of generative AI – as well as its potential pitfalls.
At Clarivate, we are incorporating generative AI tools into our Cortellis suite of health data solutions, starting with Cortellis Competitive Intelligence, Disease Landscape & Forecast and Drug Timelines and Success Rates (DTSR), to enable natural language searches across multiple structured and unstructured datasets – a major timesaver for users.
“It gives us a way to connect the ecosystem of our products and summarize findings to bring additional insights to our life science customers,” says Romeo Radman, Vice President, Product Management at Clarivate.
“We utilise large language models to help understand a query, then we run that against our databases and use summarisation to deliver clear, concise answers.
“This enables them to deliver innovative medicines and medical technologies to patients faster.”
While generative AI’s propensity to ‘hallucinate’ has bedeviled some potential applications (such as drafting legal briefs), it’s not a concern here, Radman added, as the Clarivate enhanced search tool operates within a closed system, scouring cleaned and curated proprietary databases.
Future iterations of the tool are planned to deliver generative AI insights and conversational chat technology fueled by additional datasets, adding insights into drug discovery, clinical trials, dealmaking, biomarkers and more.
New uses for old drugs
One of the most promising avenues for the use of AI and machine learning in biopharmaceuticals is for repurposing or retargeting existing drugs.
Sildenafil, initially developed as a treatment for angina, was repurposed as the erectile dysfunction treatment Viagra.
More recently, in the early months of the COVID-19 pandemic, scientists looking for effective therapeutics found that toclizumab, marketed for the treatment of rheumatoid arthritis under the brand name Actemra, reduced risk of death and time of hospitalization for certain hospitalised patients.
In one recent instance, leveraging biomarker data, Clarivate applied a proprietary network analysis workflow to support a top 20 pharma company in identifying candidates for repositioning in a neurodegenerative disorder.
The result was the identification of 23 candidate targets, out of which two were ultimately tested and validated in animal models.
Repurposing and retargeting existing drugs can bring new and effective treatments to patients while opening up new market opportunities for biopharma companies, at lower costs and with faster development cycles than starting from scratch would entail.
Applied to large datasets, machine learning can deliver predictive analytics that identify strong candidates against specific targets.
These are just a couple of the promising areas of opportunity AI and machine learning technologies are opening up for life science companies.
While some AI applications for healthcare are far from mature and the regulatory environment is changing fast, pharmas and biotechs can benefit from assessing how AI/ML might augment their near-term operations and building their capabilities to stay ahead of competitors and better deliver treatments to patients, caregivers and clinicians.
