Your question is Getting Drugs to Market. Take a moment with it on the right.
Talk me through your thinking if you like. When you're confident, submit your answer and I'll grade it like a real screen (7/10 or better passes).
How do we get the best drugs to market?
Explain how you would turn this broad objective into a practical supervised learning problem. Define an appropriate prediction target, identify useful data sources and features, choose and train a model, and design an evaluation strategy that reflects drug-development decisions. Address temporal leakage, missing data, class imbalance, interpretability, and how predictions would support portfolio prioritization. State any assumptions explicitly and describe how you would validate the approach before deployment.