Instead of relying on serendipity, PureTech uses a structured process: 1) Identify unmet need, 2) Find a promising but flawed drug with human data, 3) Define the problem that held it back, 4) Design a solution to overcome it, and 5) Test the solution. This institutionalizes the innovation cycle for value creation.
Where Lilly pursued a challenging medicinal chemistry approach to make a drug more specific, PureTech's Karuna succeeded with a simpler biological solution. They paired the drug with an existing one that blocked its effects outside the brain, mitigating side effects without altering the core, promising molecule.
Discontinued drugs aren't hard to identify; the real challenge is navigating the out-licensing process inside a large pharma company. Without an internal champion to drive the complex approvals for a non-priority asset, promising drugs can languish on the shelf due to corporate inertia, not a desire to hide them.
To combat bias, the team contractually agrees on strict, predefined success metrics for major milestones *before* any data is generated. A program either meets the criteria or it doesn't, removing ambiguity from go/no-go decisions. This discipline is applied both internally and at the board level for spun-out companies.
PureTech uses AI to accelerate the initial steps of its process: identifying promising discontinued drugs and pinpointing what held them back. However, the crucial step of devising the scientific solution to fix the drug remains a human-driven, creative insight process, blending AI's scale with human ingenuity.
By centralizing resources (hub), PureTech can dispassionately kill failing programs and reallocate talent. This structural design counters the powerful emotional and financial pressure to continue that exists when a company's survival is tied to a single drug, as people's livelihoods aren't dependent on one program's success.