iRaPCA and SOMoC: Development and Validation of Web Applications for New Approaches for the Clustering of Small Molecules

Abstract

The clustering of small molecules implies the organization of a group of chemical structures into smaller subgroups with similar features. Clustering has important applications to sample chemical datasets or libraries in a representative manner (e.g., to choose, from a virtual screening hit list, a chemically diverse subset of compounds to be submitted to experimental confirmation, or to split datasets into representative training and validation sets when implementing machine learning models). Most strategies for clustering molecules are based on molecular fingerprints and hierarchical clustering algorithms. Here, two open-source in-house methodologies for clustering of small molecules are presented: iterative Random subspace Principal Component Analysis clustering (iRaPCA), an iterative approach based on feature bagging, dimensionality reduction, and K-means optimization; and Silhouette Optimized Molecular Clustering (SOMoC), which combines molecular fingerprints with the Uniform Manifold Approximation and Projection (UMAP) and Gaussian Mixture Model algorithm (GMM). In a benchmarking exercise, the performance of both clustering methods has been examined across 29 datasets containing between 100 and 5000 small molecules, comparing these results with those given by two other well-known clustering methods, Ward and Butina. iRaPCA and SOMoC consistently showed the best performance across these 29 datasets, both in terms of within-cluster and between-cluster distances. Both iRaPCA and SOMoC have been implemented as free Web Apps and standalone applications, to allow their use to a wide audience within the scientific community.

On the Best Way to Cluster NCI-60 Molecules

“…The uniform manifold approximation and projection (UMAP) is a non-linear dimensionality reduction algorithm that seeks to learn the manifold structure of the data and find a low-dimensional embedding while preserving the essential topological structure of that manifold [ 29 ]. While UMAP has been used for dimensionality reduction [ 30 ], it has also been used for clustering [ 11 ]. UMAP has four basic parameters to control the impact on the resulting embedding.…”

“…The clustering of biological entities, such as small molecules, can be performed using different approaches, including hierarchical clustering (HC), distribution-based clustering, and density-based clustering [ 10 ]. For example, several clustering algorithms, including hierarchical, Taylor–Butina, and UMAP clustering, have been compared on 29 data sets with between 100 and 5000 small molecules [ 11 ]. In addition, hierarchical clustering has been used to cluster molecules from the PubChem database [ 12 ], and Taylor–Butina clustering has been used to cluster molecules from the MolPort database [ 13 ].…”

Identification of novel inhibitors for SARS-CoV-2 as therapeutic options using machine learning-based virtual screening, molecular docking and MD simulation

“…The dimensionality was reduced by performing the PCA. The process is based on the principle of feature bagging ( Prada Gori et al, 2022 ). The conventional feature extraction and data representation method used extensively in the fields of pattern recognition is principal component analysis (PCA), generally called as Karhunen-Loeve expansion.…”

A Combined Ligand- and Structure-Based Virtual Screening To Identify Novel NaV1.2 Blockers: In Vitro Patch Clamp Validation and In Vivo Anticonvulsant Activity

“…The molecules were clustered using Silhouette Optimized Molecular Clustering (SOMoC), an in-house clustering method. Briefly, SOMoC can be described as a sequential combination of molecular fingerprinting, dimensionality reduction using the Uniform Manifold Approximation and Projection (UMAP) algorithm, and clustering using the Gaussian Mixture Model (GMM) algorithm . Training (80%) and test (20%) partitions were then sampled in a stratified manner using the cluster assignments seeking to maintain similar distributions for the dependent variable (pIC 50 ).…”