Strategic Key Performance Indicators for AI in Lead Optimization

Area of Science:

• Drug Discovery
• Computational Chemistry
• Medicinal Chemistry

Background:

• Pharmaceutical R&D faces persistent challenges with high costs and failure rates.
• Artificial intelligence (AI) integration in R&D shows potential to enhance productivity, with AI-native companies potentially outperforming traditional ones.
• Current R&D metrics focus on technical model performance, not strategic impact, hindering AI's full potential.

Purpose of the Study:

• To address the need for strategic impact measurement in AI-driven drug discovery.
• To discuss challenges in applying Key Performance Indicators (KPIs) within the unique drug discovery environment.
• To propose a new framework of impact-oriented KPIs for AI in R&D.

Main Methods:

• Perspective piece based on discussions within the EFMC2 community.
• Analysis of the shift from expert-driven computer-aided drug design (CADD) to semi-autonomous AI.
• Development of recommendations for designing strategic KPIs.

Main Results:

• Measuring technical model performance is insufficient for assessing AI's strategic impact in drug discovery.
• The transition to AI necessitates a shift towards impact-oriented KPIs.
• Effective KPIs should drive adoption, foster innovation, and objectively assess digital tool performance.

Conclusions:

• A new framework of impact-oriented KPIs is crucial for realizing AI's potential in pharmaceutical R&D.
• Strategic KPIs will help objectively evaluate the effectiveness of digital tools in drug discovery.
• Implementing these KPIs will guide the adoption and innovation of AI in the R&D landscape.