Interview
a16z Podcast | Move Fast But Don't Break Things (When It Comes to Computational Biology)
- Pharma companies are expected to transition from fully integrated in-house operations to a leaner, outsourced "syndicator" model similar to movie studios, driven by shareholder pressure for returns amid rising development risks and costs.
- The industry will undergo a gradual shift to cloud-based services, progressing from initial pilot projects to larger adoption cycles as security concerns are resolved.
- "Cloud biology" systems utilizing robot-driven experiments are projected to become significantly cheaper and more reliable than traditional manual methods.
- A stealth company aims to digitize and automate animal model processes unchanged since the 1950s by applying sensors and big data to reduce costs and improve reliability.
- Data sharing across the industry is predicted to become inevitable as payers aggregate databases to gain insights exceeding current pharmaceutical capabilities.
- "Virtual pharma companies" specializing in outsourcing and contracting are expected to grow as a successful alternative to traditional legacy infrastructure.
- Innovation is expected to originate primarily from outside big pharma, with legacy companies succeeding only by adopting external disruptions and virtual models.
- The cultural gap between Silicon Valley and the pharmaceutical sector is anticipated to close as both sides recognize the value of computational benefits and software engineering perspectives.
- Data science and computation integration is projected to be "really fantastic" over the next 10 years, fueled by the deterministic decrease in costs for compute and genomics.
- Personalized medicine capable of generating drug formulations "on the spot" from immediate biological samples is expected to become feasible eventually as sufficient population data is accumulated.
- The pharmaceutical industry is projected to eventually reduce or eliminate reliance on animal models in favor of computer simulations, which are expected to predict human outcomes more accurately than current mouse models.
- Regulatory acceptance of computer-simulated trials instead of animal testing is expected to occur "in the decades to come," contingent upon the production of substantial hard evidence proving safety and reliability.
- The FDA is expected to accelerate acceptance of digital clinical data entry, moving away from paper-based systems as confidence in computerized record integrity grows.
- The healthcare system is expected to undergo "disintermediation" and pricing improvements driven by patient demand for on-demand services, though the transition is not expected to be easy or occur soon.