Autocorrelation descriptor improvements for QSAR: 2DA_Sign and 3DA_Sign
- 31 December 2015
- journal article
- research article
- Published by Springer Nature in Journal of Computer-Aided Molecular Design
- Vol. 30 (3) , 209-217
Abstract
Quantitative structure–activity relationship (QSAR) is a branch of computer aided drug discovery that relates chemical structures to biological activity. Two well established and related QSAR descriptors are two- and three-dimensional autocorrelation (2DA and 3DA). These descriptors encode the relative position of atoms or atom properties by calculating the separation between atom pairs in terms of number of bonds (2DA) or Euclidean distance (3DA). The sums of all values computed for a given small molecule are collected in a histogram. Atom properties can be added with a coefficient that is the product of atom properties for each pair. This procedure can lead to information loss when signed atom properties are considered such as partial charge. For example, the product of two positive charges is indistinguishable from the product of two equivalent negative charges. In this paper, we present variations of 2DA and 3DA called 2DA_Sign and 3DA_Sign that avoid information loss by splitting unique sign pairs into individual histograms. We evaluate these variations with models trained on nine datasets spanning a range of drug target classes. Both 2DA_Sign and 3DA_Sign significantly increase model performance across all datasets when compared with traditional 2DA and 3DA. Lastly, we find that limiting 3DA_Sign to maximum atom pair distances of 6 Å instead of 12 Å further increases model performance, suggesting that conformational flexibility may hinder performance with longer 3DA descriptors. Consistent with this finding, limiting the number of bonds in 2DA_Sign from 11 to 5 fails to improve performance.Keywords
Funding Information
- National Institutes of Health (GM080403, GM099842, DK097376, HL122010, GM073151, AI117905)
- National Science Foundation (1305874)
This publication has 25 references indexed in Scilit:
- Benchmarking Ligand-Based Virtual High-Throughput Screening with the PubChem DatabaseMolecules, 2013
- CAUTION: Popular “Benchmark” Data Sets Do Not Distinguish the Merits of 3D QSAR MethodsJournal of Chemical Information and Modeling, 2009
- Fast Assignment of Accurate Partial Atomic Charges: An Electronegativity Equalization Method that Accounts for Alternate Resonance FormsJournal of Chemical Information and Computer Sciences, 2003
- Deriving the 3D structure of organic molecules from their infrared spectraVibrational Spectroscopy, 1999
- Topological and Stereochemical Molecular Descriptors for Databases Useful in QSAR, Similarity/Dissimilarity and Drug DesignSAR and QSAR in Environmental Research, 1998
- Overcoming the Limitations of a Connection Table Description: A Universal Representation of Chemical SpeciesJournal of Chemical Information and Computer Sciences, 1997
- From atoms and bonds to three-dimensional atomic coordinates: automatic model buildersChemical Reviews, 1993
- Additivity methods in molecular polarizabilityJournal of the American Chemical Society, 1990
- Quantitative models of gas-phase proton-transfer reactions involving alcohols, ethers, and their thio analogs. Correlation analyses based on residual electronegativity and effective polarizabilityJournal of the American Chemical Society, 1984
- Molecular Orbital Theory for Organic ChemistsJournal of the Electrochemical Society, 1962