liveBook · Manning

Fragment-based drug design

As explained in Build AI Drug Discovery Pipelines

Fragment-based drug discovery, or fragment-based design, identifies small molecular fragments that bind a target and elaborates or links them into lead compounds.

Fragment-based drug design screens a library of small molecular fragments for target binding, then elaborates successful fragments into stronger lead compounds. Unlike conventional screening, fragments probe binding-site structure extensively despite using a smaller library. section 7.3 section 7.3.1

A practical fragment-based workflow includes:

  • Screen a fragment library and identify multiple fragments with measurable affinity for the target. section 7.3
  • Grow one fragment by adding R-groups; rigid additions are generally preferred because flexible groups can make affinity and interpretation less predictable. section 7.3.1
  • Link two non-competitive fragments occupying different sub-pockets, while computing a linker orientation that balances connection and conformational freedom. section 7.3.1
  • Merge fragments that partially occupy the same region into one structure combining useful portions of both. section 7.3.1
  • Optimize the expanded compound and verify that its binding profile improves over the individual fragments; many resulting compounds will not become suitable leads. section 7.3.1

Fragments are weak binders, and measured binding may reflect a crystallized, static protein structure. Consequently, expansion, linking, or merging does not guarantee improved affinity or a viable lead. section 7.3.1

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Explains screening and three elaboration strategies, but defers comprehensive coverage of additional strategies and later optimization complexity.
Examples
Uses fragment libraries, target sub-pockets, R-groups, linkers, and merged fragments as concrete workflow examples.
What you can do after
Apply fragment growing, linking, and merging to convert binding fragments into lead candidates for further evaluation.
How it connects
Connects fragment screening with unsupervised learning, pharmacophore modeling, density estimation, and lead optimization.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Build AI Drug Discovery Pipelines. The section links point to the parts of the book it draws on.