Take the screening strategy challenge

For a single potential source of failure (hazard), with two independent methods that give yes/no answers:

• How many choices of screening strategy are there (we think there are no less than six)?
• How can you sensibly combine two successive methods, such as a low-cost in silico method and a more expensive but more accurate in vitro screen):
  • progressing only the compounds that pass both tests (‘double filter’ strategy)?
  • failing only the compounds that throw up alerts in both tests (‘sentinel’ strategy)?
• What is the right choice of screening strategy for a given project, given the performance of the predictive and screening methods on the relevant target and chemotypes?

We think that the demonstration below, which requires Flash Player 10, supports deeper insight into these and related questions. As well as simply calculating impact of the different strategies on pipeline value, you can explore how the strategy value depends on predictive and screening method performance and on the project context (risk, cost of failure and value of success).

Screening strategy explorer demonstration

Many find it hard to get up to 90% of the value that could be delivered by using optimal plans, given the complexity of even this somewhat simplified planning problem (we recognize the Real World Complications). Even if your score is 95% of optimal, the potential improvement is no less than 5% of total pipeline value, for minimal investment in training and support

We have recently developed versions of this tool that support critical thinking about more tactical R&D decisions, such as choice of cutoff for progression (including decisions involving biomarker read-outs). Contact us for more details and to discuss your interests in R&D planning and how to get the most value from predictive technologies.