Taylor-Butina clustering selects chemically central compounds from a large molecular library, then assigns nearby compounds using a fingerprint-distance cutoff. It is useful when exhaustive clustering is too slow for millions of compounds. section 7.2.3
Taylor-Butina clustering
Taylor-Butina clustering is a fingerprint-based clustering method that selects high-neighbor compounds as centroids and assigns nearby compounds using an exclusion threshold.
How Build AI Drug Discovery Pipelines covers it
- How deep it goes
- Explains pairwise-distance construction, cutoff variation, complexity, and memory tradeoffs, but not optimized implementation details.
- Examples
- Uses RDKit fingerprints, cutoff sweeps, cluster-size distributions, silhouette scores, and redundancy calculations.
- What it assumes
- Assumes molecular fingerprints, pairwise similarity, and basic clustering terminology; demonstrates the computational workflow explicitly.
- What you can do after
- Apply it via RDKit and Butina to cluster fingerprints and evaluate diverse compound samples.
- How it connects
- Connects molecular fingerprints, Tanimoto similarity, clustering cutoffs, dimensionality reduction, and library-diversity evaluation.
Related concepts
- clusteringkind of · Clustering is the process of grouping data items so that items within a group are more similar to one another than to items in other groups.
- K-means clusteringalternative to · K-means clustering is an algorithm that partitions data into a specified number of clusters by iteratively assigning observations to the nearest centroid and updating each centroid to the mean of its assigned observations.
- Jaccard similarityuses · Jaccard similarity is a measure of overlap between two sets, calculated as the size of their intersection divided by the size of their union.
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.
