SmartDrugDiscovery
ExploreCollaboratePromote
Sign in
SmartDrugDiscovery

Precision in research — explore, collaborate, and promote across drug discovery.

AboutExploreCollaboratePromoteContact
AllPapersDatasetsToolsNewsPodcast
arrow_backAll episodes
EP 53

The Math of AI Drug Discovery

This episode decodes the math behind the hype of AI drug discovery, exploring how companies like Insilico Medicine are using AI to develop new drugs and the financial models that drive their valuation. The discussion delves into the complexities of biotech valuation, including the role of risk-adjusted net present value and the challenges of navigating the 'valley of death' in clinical trials.

AI drug discoverybiotech valuationrisk-adjusted net present valueclinical trialsbusiness models
play_arrowListen to episode

Key Concepts

Biotech Valuation

  • The biotech industry operates under extreme information asymmetry.
  • The 'valuation gap' exists due to the difference in knowledge between the management team and external investors.
  • The risk-adjusted net present value (RNPV) framework is the industry gold standard for pre-revenue biotech valuation.

Risk-Adjusted Net Present Value (RNPV)

  • RNPV adjusts future expected revenue based on the mathematical probability of success at each clinical phase.
  • The framework takes into account the probability of success, the cost of capital, and the time value of money.

Clinical Trials

  • The 'valley of death' in phase two clinical trials has a historically low transition probability of 35-40%.
  • Phase two trials are the first time a drug is introduced into patients who actually have the disease, making it a critical phase for efficacy testing.

Business Models in AI Drug Discovery

  • AI drug discovery companies are adopting various business models to navigate the risk of clinical failure.
  • The models include the pure software or platform model, the AI-enabled CRO model, the partnership-led platform biotech model, and the AI pharma model.

Real Options Valuation (ROV)

  • ROV is a framework that views clinical trials as a series of stage choices or options.
  • High volatility increases the value of the company in the ROV framework.

Episode Summary

  • check_circleThe biotech industry operates under extreme information asymmetry, making valuation challenging.
  • check_circleThe 'valuation gap' exists due to the difference in knowledge between the management team and external investors.
  • check_circleThe risk-adjusted net present value (RNPV) framework is the industry gold standard for pre-revenue biotech valuation.
  • check_circleRNPV adjusts future expected revenue based on the mathematical probability of success at each clinical phase.
  • check_circleThe 'valley of death' in phase two clinical trials has a historically low transition probability of 35-40%.
  • check_circleAI drug discovery companies are adopting various business models to navigate the risk of clinical failure.
  • check_circleThe 'sum-of-the-parts' methodology is used to evaluate the market cap of hybrid companies by breaking them down into separate components.
  • check_circleReal options valuation (ROV) is a framework that views clinical trials as a series of stage choices or options, where high volatility increases the value of the company.
  • check_circleThe patent cliff is a hard mathematical stop on any real options valuation model, as it marks the end of a drug's revenue cycle.
  • check_circleThe rise of cell and gene therapy (CGT) threatens to break the entire mathematical framework of RNPV models.

Full Transcript

Discussion

Join the discussion — sign in to leave a comment.

Log in to comment
biotech

Live Literature

Current papers related to this episode's topics.

podcasts

Related Episodes