Skip to content

Chapter 14 of Generative

Generative design loops soon

Closing the loop: multi-objective optimisation, docking-guided generation, active learning and an introduction to protein design.

soon This chapter is planned and not yet written. The outline below is the current plan; it may change.

What it will cover

  • Multi-objective optimisation and Pareto fronts
  • Scoring functions, property predictors and docking in the loop
  • Active learning and Bayesian optimisation for experiments
  • Protein design concepts
  • Pitfalls: reward hacking, synthesisability, leakage into the scorer
  • Lab 14: a design loop that optimises solubility and a docking score

Planned learning outcomes

  1. Set up a generate–score–select loop with several objectives.
  2. Use active learning to choose the next experiments under a budget.
  3. Recognise and guard against the failure modes of optimisation against a learned scorer.

In the meantime

Part IV assumes Parts I–III. Start with Chapter 0 if you are new.

Edit this pageReport a problemLast updated 2026-11-02