From The Editor | July 6, 2017

Takeda Uses AI To Identify Promising Molecules

Ed Miseta

By Ed Miseta, Chief Editor, Clinical Leader
Follow Me On Twitter @EdClinical

Takeda Uses AI To Identify Promising Molecules

Efficiently and effectively identifying high-quality drug candidates for clinical development is challenging, even for a large company like Takeda that has extensive R&D expertise and therapeutic experience in the areas of oncology, gastroenterology, and diseases of the central nervous system. To combat this challenge, Takeda, not unlike others in the industry, is looking to augment its internal drug discovery abilities with innovative capabilities from outside the company – at pharma and bio companies, as well as academic institutions around the world.

Manage Large Amounts Of Data

The search for partners with unique drug discovery capabilities prompted Takeda to enter into an alliance with Numerate. Numerate is a software company that has developed algorithms that analyze large amounts of data and find promising unexplored chemical space for the human drug targets Takeda wants to pursue.

“They [Numerate] are able to leverage publicly available data as well as our proprietary data to computationally envision new chemical space for novel therapies,” says David Weitz, Head of Takeda California and Global Research Externalization for Takeda California. “In addition, Numerate has algorithms that help it envision how to engineer out compound liabilities during lead optimization to ultimately arrive at high-quality drug candidates.”

Individuals can look at large amounts of data and perform modeling techniques, but there are limits to what the human mind can process. What makes Numerate’s technology different is that it can peer across large data sets and identify opportunities that might otherwise be overlooked. And in addition to finding new chemical space, Numerate’s algorithms enable the company to infer how to refine compounds to improve desired drug properties (as well as engineer out undesired properties).

“Initially, you try to discover chemical starting points to target a given protein,” notes Weitz. “Later, you focus on how to refine those compounds into high-quality candidates with the desired drug properties for the disease being targeted.  Being able to do this efficiently and with insights drawn from large data sets is the promise of Numerate’s AI platform.”

An Alliance To Find Exceptional Candidates

One of the key features of this partnership is that it’s a true drug discovery alliance that seeks to wed the therapeutic insights and R&D capabilities of Takeda with Numerate’s unique capabilities.  The two companies agree on therapeutic targets that are of interest to Takeda and are targets that Numerate feels it can succeed with using its AI platform and the available data. To achieve speed and agility, and as a sign of mutual trust, the parties agreed that Numerate will work largely independently with Takeda playing more of an advisory and supporting role during the discovery phase.  Once promising clinical candidates are identified, Takeda will do what it does best – develop and commercialize the candidates into new therapies that benefit patients.  Typically, the candidates Takeda receives under the alliance will take about one year to reach the clinic.

“Our hope is that the AI platform will enable Numerate, via its ability to analyze large volumes of data using computational power, to arrive at high-quality compounds for novel therapeutic approaches that have the potential to satisfy significant unmet needs for patients,” states Weitz. “We know that patients are waiting and are eager to discover these candidates with Numerate so Takeda can get them to the clinic and ultimately to patients.”

Takeda’s overriding goal is to identify therapy opportunities that will be the most transformative for patients, and which fall into its therapeutic areas of interest.  As an outward-looking R&D organization, Takeda is always searching for novel technologies that can help it quickly get to quality new medicines. That philosophy led it to align with Numerate.  “We cannot perform all of the discovery work internally nor would we want to,” adds Weitz. “Innovation is all around us.  We actively try to identify companies and academic institutions that can bring pieces of the puzzle that are complementary and synergistic to our own R&D repertoire so that together we can deliver new therapeutics that meaningfully impact patient lives.”