Pre-Conference Day

As modelling capabilities continue to expand, many organisations are now operating multiple modelling approaches in parallel, including pharmacometrics, QSP, machine learning, and AI-enabled analytics. Yet the greatest value is often realised not from any individual model, but from understanding how these approaches can work together to answer development questions more effectively. This interactive workshop explores how organisations are building integrated modelling strategies that connect mechanistic understanding, quantitative prediction, and data-driven insights to support decisions across discovery and development.

This workshop will gather experts to discuss:

• Understand where pharmacometrics, QSP, machine learning, and AI each provide unique value across the development lifecycle, and where overlap can create duplication or inefficiency
• Explore practical frameworks for connecting modelling approaches, including how outputs from one model can inform, refine, or validate another
• Examine how AI and machine learning can be used to incorporate complex data sources such as imaging, genomics, transcriptomics, proteomics, and real-world datasets into modelling workflows that were previously constrained by data complexity
• Review case studies demonstrating how integrated modelling strategies improved dose selection, patient stratification, biomarker identification, and clinical trial design
• Discuss the challenges of transparency, validation, and regulatory acceptance when combining mechanistic and AI-enabled approaches

Regulatory acceptance of MIDD evidence is the goal, but it is rarely where the resistance starts. Before a model reaches a submission, it has to survive scrutiny from clinical scientists who distrust mechanistic outputs, biostatisticians who speak a different statistical language, and senior leaders weighing up whether to back a modelling strategy they cannot directly evaluate. Internal buy-in is not a soft problem: it is what determines whether a model is stress-tested enough to hold up when regulators push back. This workshop addresses both sides of the credibility challenge, building the internal case for MIDD and QSP outputs alongside the regulatory one, because the two are more connected than most teams treat them.

This workshop will gather experts to discuss:

• Examine what a credible pre-IND MIDD package looks like across different modalities, including small molecules, ADCs, and biologics, and where the bar is currently set by FDA and EMA
• Discuss the growing role of New Approach Methodologies (NAMs) and in vitro data in supplementing or replacing animal studies, and how to integrate these into your modelling framework
• Work through case examples of pre-clinical MIDD submissions that succeeded or fell short, identifying the decisions, datasets, and documentation that made the difference