Join this workshop to move beyond the “AI buzzword” and confront the biological hurdles of antigen heterogeneity. As we pivot toward more complex engineering, this session addresses the critical uncertainty of target density: determining whether 20% or 80% cellular expression is the threshold for clinical success.

• Using machine learning to synthesize datasets to profile antigen prevalence across primary vs. metastatic tissues
• Identifying which novel targets are tumor-specific versus those with risky low-level expression in healthy tissues
• Deep-diving into the antigen expression level needed to trigger the action of a T-cell engager
• Developing AI algorithms that move beyond simple “positive/ negative” IHC scores to more predictive “probability of response” scores based on complex expression patterns
• Discussing how clinical trial failures can be fed back into AI models to refine the next generation of discovery