Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2014 Dec 11;6(1):48.
doi: 10.1186/s13321-014-0048-0. eCollection 2014.

Prediction-driven matched molecular pairs to interpret QSARs and aid the molecular optimization process

Yurii Sushko  1 , Sergii Novotarskyi  1 , Robert Körner  1 , Joachim Vogt  1 , Ahmed Abdelaziz  1 , Igor V Tetko  2
Affiliations

Prediction-driven matched molecular pairs to interpret QSARs and aid the molecular optimization process

Yurii Sushko et al. J Cheminform. .

Abstract

Background: QSAR is an established and powerful method for cheap in silico assessment of physicochemical properties and biological activities of chemical compounds. However, QSAR models are rather complex mathematical constructs that cannot easily be interpreted. Medicinal chemists would benefit from practical guidance regarding which molecules to synthesize. Another possible approach is analysis of pairs of very similar molecules, so-called matched molecular pairs (MMPs). Such an approach allows identification of molecular transformations that affect particular activities (e.g. toxicity). In contrast to QSAR, chemical interpretation of these transformations is straightforward. Furthermore, such transformations can give medicinal chemists useful hints for the hit-to-lead optimization process.

Results: The current study suggests a combination of QSAR and MMP approaches by finding MMP transformations based on QSAR predictions for large chemical datasets. The study shows that such an approach, referred to as prediction-driven MMP analysis, is a useful tool for medicinal chemists, allowing identification of large numbers of "interesting" transformations that can be used to drive the molecular optimization process. All the methodological developments have been implemented as software products available online as part of OCHEM (http://ochem.eu/).

Conclusions: The prediction-driven MMPs methodology was exemplified by two use cases: modelling of aquatic toxicity and CYP3A4 inhibition. This approach helped us to interpret QSAR models and allowed identification of a number of "significant" molecular transformations that affect the desired properties. This can facilitate drug design as a part of molecular optimization process. Graphical AbstractMolecular matched pairs and transformation graphs facilitate interpretable molecular optimisation process.

Keywords: Interpretation; Inverse QSAR; MMP; Matched molecular pairs; Medicinal chemistry; Molecule optimization; OCHEM; Online chemical modelling environment; QSAR.

PubMed Disclaimer

Figures

Graphical Abstract
Graphical Abstract
Molecular matched pairs and transformation graphs facilitate interpretable molecular optimisation process.
Figure 1
Figure 1
Four typical examples of matched molecular pairs (MMPs).
Figure 2
Figure 2
Activity cliff example. A molecule inactive according to the Ames test becomes active after a minor structural change. Activity cliffs represent interesting cases for activity interpretation.
Figure 3
Figure 3
Molecules transformed within the scope of this study. All eight molecules were selected from the DrugBank database.
Figure 4
Figure 4
Permethrin optimization examples. Six exemplary modifications of Permethrin that significantly decrease its predicted aquatic toxicity (growth inhibition concentration). A decrease of 1–2 log units can be achieved by making only minor structural changes.
Figure 5
Figure 5
Transformation graph for Permethrin optimization. Graph of the transformations that affect the aquatic toxicity of the Permethrin molecule. The graph includes 393 transformations that provide replacements of several structural groups by less toxic variants. A cluster with replacements of ether groups is shown in detail together with a few examples of optimized molecules.
Figure 6
Figure 6
Experimental and predicted evidence supporting the toxicity-reducing effect of a selected transformation.
Figure 7
Figure 7
Toxicity optimization: statistically and practically significant transformations. The chart shows interesting transformations that are both statistically significant (significance level >2, p-value <0.01) and effective (mean toxicity change at least one log unit). A number of transformations that did not have sufficient measured pairs became significant when combined with predicted pairs (were “amplified”, shown as solid red circles).
Figure 8
Figure 8
Transformations graph of CYP3A4 optimization of Hexestrol.
Figure 9
Figure 9
Sample modified molecules obtained from Hexestrol after CYP3A4 inhibition optimization. Overdestructive changes can be avoided by additional filtering by structure similarity (e.g. Tanimoto similarity).
Figure 10
Figure 10
Experimental and predicted evidence supporting the CYP inhibition-reducing effect of a selected transformation.
Figure 11
Figure 11
CYP inhibition optimization: statistically and practically significant transformations. The chart shows interesting transformations that are both statistically significant (significance level >2, p-value <0.01) and effective (ratio of deactivated molecules at least 55%). A number of transformations that did not have sufficient measured pairs became significant when combined with predicted pairs (were “amplified”, shown as solid red circles).
Figure 12
Figure 12
Importance of chemical context for transformations. The same transformation can have significantly different effects in different contexts.
Figure 13
Figure 13
Exemplary molecular transformations. Single and double-point transformations shown.
Figure 14
Figure 14
Effect of a transformation on molecular properties. A) A simple transformation and the distribution of its effect on the octanol/water partition coefficient. The histogram is visually biased to positive values: on average, this transformation increases lipophilicity. B) Replacement of carbon by bromine significantly increases aquatic toxicity.
Figure 15
Figure 15
A transformations graph for aquatic toxicity. Arrows point towards the direction of increasing toxicity. For example, it can be seen that the presence of bromine is potentially more toxic than the presence of chlorine, whereas the hydroxyl group is the least toxic residual in this example.
Figure 16
Figure 16
A delta-pair chart for an aquatic toxicity model. Three representative cases of activity cliffs are shown. The right part shows the significant transformations for aquatic toxicity.
Figure 17
Figure 17
A simplified database schema to store MMPs, transformations and transformation annotations.
See this image and copyright information in PMC

References

    1. Ekins S, Waller CL, Swaan PW, Cruciani G, Wrighton SA, Wikel JH. Progress in predicting human ADME parameters in silico. J Pharmacol Toxicol Methods. 2000;44(1):251–272. doi: 10.1016/S1056-8719(00)00109-X. - DOI - PubMed
    1. Perkins R, Fang H, Tong W, Welsh WJ. Quantitative structure-activity relationship methods: perspectives on drug discovery and toxicology. Environ Toxicol Chem. 2003;22(8):1666–1679. doi: 10.1897/01-171. - DOI - PubMed
    1. Verma J, Khedkar VM, Coutinho EC. 3D-QSAR in drug design–a review. Curr Top Med Chem. 2010;10(1):95–115. doi: 10.2174/156802610790232260. - DOI - PubMed
    1. OECD principles for the validation, for regulatory purposes, of (quantitative) structure-activity relationship models. [http://www.oecd.org/env/ehs/risk-assessment/validationofqsarmodels.htm]
    1. Griffen E, Leach AG, Robb GR, Warner DJ. Matched molecular pairs as a medicinal chemistry tool. J Med Chem. 2011;54(22):7739–7750. doi: 10.1021/jm200452d. - DOI - PubMed

LinkOut - more resources