liveBook · Manning

Pharmacophore-based screening

As explained in Build AI Drug Discovery Pipelines

Pharmacophore-based sampling selects compounds that match hypothesized spatial and chemical features required for target binding.

Pharmacophore-based screening selects compounds whose spatial arrangements and chemical features match a binding hypothesis. It is useful when several active compounds are known, or when structural information about a target binding site supports a pharmacophore hypothesis. section 2.1.1 section 9.2.4

A ligand-based pharmacophore search typically proceeds by:

  • Generating possible three-dimensional configurations for known active compounds. section 2.1.1
  • Aligning configurations to identify structural features shared by the active compounds. section 2.1.1
  • Scoring features according to how commonly they occur. section 2.1.1
  • Screening unknown compounds for high-scoring feature matches. section 2.1.1

The method belongs to ligand-based virtual screening when the target’s three-dimensional structure is unavailable. With target binding-site information, pharmacophore hypotheses can instead guide the initial sample for docking or active-learning workflows. section 2.1.1 section 9.2.4

Figure 2.1 Virtual screening pipelines are often hierarchical, proceeding with a broad, large number of compounds through simpler, less restrictive, and computationally cheaper processes to whittle down the quantity of compounds. If the 3D structure of the target is not defined, we’re restricted to ligand-based virtual screening. If it is defined, then we can explore structure-based virtual screening, potentially as a downstream component to ligand-based virtual screening.

Use the virtual-screening pipeline diagram to place pharmacophore filtering among hierarchical screening stages that progressively reduce a compound library. section 2.1.1

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Explains configuration generation, alignment, feature scoring, and library searching, but does not detail pharmacophore feature representations or software implementation.
Examples
Uses several known active compounds, their possible three-dimensional configurations, and unknown library compounds as screening examples.
What it assumes
Assumes multiple known active compounds or target binding-site information, while distinguishing search-based screening from machine learning requiring active and inactive examples.
What you can do after
Apply it via a pharmacophore search to prioritize unknown compounds for further virtual screening or expensive docking simulations.
How it connects
Connects pharmacophore searching with virtual screening, similarity searching, docking, active learning, and ligand-based machine learning.

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.