liveBook · Manning

Homology modeling

As explained in Build AI Drug Discovery Pipelines

Homology modeling predicts a protein structure by using experimentally determined structures of evolutionarily related proteins as templates.

Homology modeling predicts a target protein’s structure by searching databases of known structures for proteins with similar sequences and using those relatives as templates. The method relies on evolution having produced structurally informative examples already, making it useful when close evolutionary relatives exist. section 12.1.2

Use homology modeling as a practical structure-prediction workflow with a clear limitation:

  • Search known-structure databases for proteins related to the target sequence. section 12.1.2
  • Build a target structure from the selected evolutionary relatives rather than simulating every atomic motion or exhaustively sampling conformations. section 12.1.2
  • Treat the result as uncertain when templates are weak or absent: the approach works well for close relatives but fails for novel folds. section 12.1.2

Target structure for rentosertib discovery

A rentosertib program used homology modeling because an experimental structure for the target was unavailable. The resulting structural uncertainty affected docking calculations, so iterative experimental testing was used to address it. section 13.3.1

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Explains database searching, template use, and failure on novel folds, but stops short of alignment or model-building implementation details.
Examples
Uses known-structure databases and a rentosertib target model to illustrate template selection and uncertainty in docking.
What you can do after
Evaluate when template-based structure prediction can support docking and when experimental testing is needed to address structural uncertainty.
How it connects
Connects homology modeling with protein structure prediction, molecular dynamics, deep learning, docking calculations, and experimental validation.

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.