Pharmacy Daily | AI can act as a new drugs accelerator

GRANT Case, Regional VP, APAC at AI company Dataiku explains how pharmaceutical and life sciences companies can accelerate innovation through AI?

Do you have an opinion to share? Let us know in up to 400 words via email to feedback@pharmacydaily.com.au.

Drug development is a complex and costly process, with it taking anywhere from 10 to 15 years and around US$1 billion (A$1.47 billion)to develop a single successful drug.

Despite the hefty investment, only around 10% of drug development projects eventually get approved.

As pharmaceutical companies continue to digitise all aspects of the business and build up their curated knowledge bases, a new way to connect data sources has come to the fore.

Through the use of artificial intelligence (AI), machine learning (ML) and graph technology, pharmaceutical companies can now explore the complex, interrelated relationships between drugs, diseases, and genes to unlock new ways to develop new indications for lifesaving drugs for patients.

Historically, drug repurposing involved traditional structured data mapping and manual experimentation from bench scientists and biologists.

This process often proved to be time-consuming, costly and mistake-riddled.

Though the use of AI in pharmaceuticals is still in its infancy, drug repurposing was one of the early applications that delivered strong results from the outset.

In order to leverage this technology, companies must first draw data from internal and external sources, such as drug structures, gene targets, pathway perturbations, disease symptoms, and more.

Using a modern data science platform will allow businesses to break down data silos and improve collaboration between business units to streamline the process of data consolidation and allow for insights to be drawn in order to identify new use cases for previously approved drugs.

This process drastically reduces timelines for new indications and saves businesses up to 80% of the costs associated with developing a brand-new drug.

Reusing existing compounds also allows these companies to improve market introduction by up to 150% as compared to a novel drug.

Pharmaceutical businesses wondering how they can undertake drug repurposing should approach it through an Everyday AI lens.

Everyday AI refers to AI that is so ingrained and intertwined with the workings of the day-to-day that it's just part of the business, not just used and developed by one team.

This approach starts with accelerating digitalisation, such as the use of electronic lab notebooks.

From there, scientists should move the results of bench experiments off of laptops and make use of modern data science platforms to allow for improved access across the business.

Such platforms empower scientists to explore historical experiment trends and mine chemical compound databases with ML in order to prioritise experiments.

Once the initial discovery and development of a new compound are complete and scientists have optimised the molecular targets, the next step is ensuring that patient populations have the right outcomes.

It's increasingly common to see biomarkers in drug labels as pharmaceutical businesses move continually to precision medicine, in order to provide patient treatment guidance.

Having the capability to visualise and comprehend all the data and associating it with patients' sub-population trends, enterprises could find that the drug has a biological mechanism of action that may not exist or may not be visible in a specific population.

From there, businesses must go back to the drawing board to re-tailor medication so that they deliver similar outcomes in different groups who have different biological expressions.

As this article has shown, drug repurposing is far from a simple exercise.

The entire process mandates meticulous investigation but can eventually act as a potent accelerator to deliver new drugs to the market in an equitable and comprehensive manner.