Trust, But Verify: On the Importance of Chemical Structure Curation in Cheminformatics and QSAR Modeling Research

By Denis Fourches, Eugene N. Muratov, Alexander Tropsha. 2010

733 citations recorded by the source corpus.

Subfields: Computational Theory and Mathematics, Materials Chemistry, Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computational Theory and Mathematics

In Materials Chemistry

In Molecular Biology

Citations in the corpus

Interpreting linear support vector machine models with heat map molecule coloring (2011, methodology): “All data sets were prepared according to the guidelines by Fourches et al. [19].”
Standardizing Benchmark Dose Calculations to Improve Science-Based Decisions in Human Health Assessments (2014, methodology): “A rigorous chemical structure curation protocol (Fourches et al. 2010) was applied to ensure that the chemical structures were standardized and that mixtures and chemicals for which descriptors cannot be calculated (i.e., inorganics, organometallics) were removed.”
Machine learning for molecular and materials science (2018, neutral): “16 The training of an ML model may be supervised, semi-supervised or unsupervised , depending upon the type and amount of available data.”
Tuning hERG out: Antitarget QSAR Models for Drug Development (2014, background): “Nevertheless, no need exists of the overabundance of models in the consensus ensemble [94].”
Comparison of Deep Learning With Multiple Machine Learning Methods and Metrics Using Diverse Drug Discovery Data Sets. (2017, background): “For all the computational modeling approaches which we may want to consider they are dependent upon the model applicability domain 112 and are affected by the quality of the underlying data, 113, 114 which may in turn determine the utility and relevance of any model and prediction 115 .”
QSAR-Based Virtual Screening: Advances and Applications in Drug Discovery (2018, background): “Considering these limitations, Fourches et al. (2010; 2015; 2016) developed the guidelines for chemical and biological data curation as a ﬁrst and mandatory step of the predictive QSAR modeling.”
Ambiguity of non-systematic chemical identifiers within and between small-molecule databases (2015, background): “The correctness of the chemical identifiers that link to the chemical structures in the databases can greatly affect the results of cheminformatics analyses [11, 12].”
UniChem: extension of InChI-based compound mapping to salt, connectivity and stereochemistry layers (2014, background): “Such debates are common [24,25] for example, and since we suspect that the growth of chemistry databases will continue to outstrip the resources to curate them, more automated mechanisms for identifying incorrectly represented molecules would be useful.”
Deep reinforcement learning for de novo drug design (2017, methodology): “(76) Briefly, explicit hydrogens were added, and specifics chemotypes such as aromatic and nitro groups were normalized using ChemAxon Standardizer.”
Exploiting drug-disease relationships for computational drug repositioning (2011, background): “many structures [20] and other chemical properties of known drug compounds contain errors, or are withheld as proprietary information.”