Clustering files of chemical structures using the Székely–Rizzo generalization of Ward's method

Abstract

Ward's method is extensively used for clustering chemical structures represented by 2D fingerprints. This paper compares Ward clusterings of 14 datasets (containing between 278 and 4332 molecules) with those obtained using the Székely-Rizzo clustering method, a generalization of Ward's method. The clusters resulting from these two methods were evaluated by the extent to which the various classifications were able to group active molecules together, using a novel criterion of clustering effectiveness. Analysis of a total of 1400 classifications (Ward and Székely-Rizzo clustering methods, fourteen different datasets, five different fingerprints and ten different distance coefficients) demonstrated the general superiority of the Székely-Rizzo method. The distance coefficient first described by Soergel performed extremely well in these experiments, and this was also the case when it was used in simulated virtual screening experiments.

Combination Rules for Group Fusion in Similarity‐Based Virtual Screening

“…[34] The additional searches were carried out using: the same BIN but with the standard belief function of Chen et al (there denoted by STD); the same BIN but with the language modelling smoothing function of Chen et al (there denoted by SMO); the BIN and belief function described by Abdo and Salim; [35] a conventional Tanimoto coefficient using binary fingerprints; a Tanimoto coefficient using fingerprints containing counts of the frequencies with which each bit occurs; a Tanimoto coefficient using fingerprints containing the square roots of the frequencies with which each bit occurs, as suggested by Arif et al; [41] and the Soergel coefficient using fingerprints containing counts of the frequencies with which each bit occurs, as suggested by Varin et al [42] The relative performance of the systems varied somewhat from dataset to dataset, but with the Soergelbased system best overall. There was no such variation in the relative performance of the fusion rules, with rCombRKP giving the best results for all four datasets, as exemplified by the results listed in Table 7 for DS1.…”

Using an unsupervised approach of Probabilistic Neural Network (PNN) for land use classification from multitemporal satellite images

“…To do so, many criteria have been described like partitioning methods, hierarchical clustering, etc. One of the most widespread hierarchical clustering methods is the Ward's method [56][57][58][59][60][61][62][63][64]. According to Hands and Everitt [64], this method achieves good results than other hierarchical methods (single-link, complete linkage, median, average linkage, etc.)…”

Information Theory and Voting Based Consensus Clustering for Combining Multiple Clusterings of Chemical Structures

“…The high value occurs when the actives are clustered tightly together and separated from the inactive molecules. The QPI is defined as: [8] QPI…”