liveBook · Manning

Prototype-based clustering

As explained in Build AI Drug Discovery Pipelines

Prototype-based clustering represents each cluster by a prototype and assigns instances according to their proximity to those prototypes.

Prototype-based clustering represents each compound cluster with a prototype, such as a compound or centroid, then assigns compounds to their nearest prototype. The standard k-means workflow initializes centroids, assigns compounds, recomputes means, and repeats until convergence. section 7.2.3

Apply the method carefully to molecular representations:

  • Avoid default k-means for binary fingerprints with Tanimoto similarity; it is better suited to normalized continuous descriptors and Euclidean distance. section 7.2.3
  • Consider Taylor-Butina clustering for cheminformatics applications, varying neighbor cutoffs and evaluating cluster quality. section 7.2.3

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Explains prototype assignment and k-means iterations, then stops short of detailed Taylor-Butina treatment.
Examples
Demonstrates centroid-based k-means and Python Taylor-Butina clustering across multiple distance cutoffs.
What it assumes
Assumes molecular fingerprints, similarity, distance, and clustering terminology are available for practical use.
What you can do after
Apply it via RDKit and evaluate clustering across cutoff values using standard clustering metrics.
How it connects
Connects prototypes with k-means, molecular fingerprints, Tanimoto similarity, and Taylor-Butina clustering.

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.