Search

Macrocycles are a promising class of compounds as therapeutics for difficult drug targets due to a favorable combination of properties: They often exhibit improved binding affinity compared to their linear counterparts due to their reduced conformational flexibility, while still being able to adapt to environments of different polarity. To assist in the rational design of macrocyclic drugs, there is need for computational methods that can accurately predict conformational ensembles of macrocycles in different environments. Molecular dynamics (MD) simulations remain one of the most accurate methods to predict ensembles quantitatively, although the accuracy is governed by the underlying force field. In this work, we benchmark four different force fields for their application to macrocycles by performing replica exchange with solute tempering (REST2) simulations of 11 macrocyclic compounds and comparing the obtained conformational ensembles to nuclear Overhauser effect (NOE) upper distance bounds from NMR experiments. Especially, the modern force fields OpenFF 2.0 and XFF yield good results, outperforming force fields like GAFF2 and OPLS/AA. We conclude that REST2 in combination with modern force fields can often produce accurate ensembles of macrocyclic compounds. However, we also highlight examples for which all examined force fields fail to produce ensembles that fulfill the experimental constraints.

Molecular flexibility is a commonly used, but not easily quantified term. It is at the core of understanding composition and size of a conformational ensemble and contributes to many molecular properties. For many computational workflows, it is necessary to reduce a conformational ensemble to meaningful representatives, however defining them and guaranteeing the ensemble’s completeness is difficult. We introduce the concepts of torsion angular bin strings (TABS) as a discrete vector representation of a conformer’s dihedral angles and the number of possible TABS (nTABS) as an estimation for the ensemble size of a molecule, respectively. Here, we show that nTABS corresponds to an upper limit for the size of the conformational space of small molecules and compare the classification of conformer ensembles by TABS with classifications by RMSD. Overcoming known drawbacks like the molecular size dependency and threshold picking of the RMSD measure, TABS is shown to meaningfully discretize the conformational space and hence allows e.g. for fast checks of the coverage of the conformational space. The current proof-of-concept implementation is based on the ETKDGv3 conformer generator as implemented in the RDKit and known torsion preferences extracted from small-molecule crystallographic data.

The concept of data fusion - the combination of information from different sources describing the same object with the expectation to generate a more accurate representation - has found application in a very broad range of disciplines. In the context of ligand-based virtual screening (VS), data fusion has been applied to combine knowledge from either different active molecules or different fingerprints to improve similarity search performance. Machine-learning (ML) methods based on fusion of multiple homogeneous classifiers, in particular random forests, have also been widely applied in the ML literature. The heterogeneous version of classifier fusion - fusing the predictions from different model types - has been less explored. Here, we investigate heterogeneous classifier fusion for ligand-based VS using three different ML methods, RF, naı̈ve Bayes (NB), and logistic regression (LR), with four 2D fingerprints, atom pairs, topological torsions, RDKit fingerprint, and circular fingerprint. The methods are compared using a previously developed benchmarking platform for 2D fingerprints which is extended to ML methods in this article. The original data sets are filtered for difficulty, and a new set of challenging data sets from ChEMBL is added. Data sets were also generated for a second use case: starting from a small set of related actives instead of diverse actives. The final fused model consistently outperforms the other approaches across the broad variety of targets studied, indicating that heterogeneous classifier fusion is a very promising approach for ligand-based VS. The new data sets together with the adapted source code for ML methods are provided in the Supporting Information.

As part of the ongoing quest to find or construct large data sets for use in validating new machine learning (ML) approaches for bioactivity prediction, it has become distressingly common for researchers to combine literature IC50data generated using different assays into a single data set. It is well-known that there are many situations where this is a scientifically risky thing to do, even when the assays are against exactly the same target, but the risks of assays being incompatible are even higher when pulling data from large collections of literature data like ChEMBL. Here, we estimate the amount of noise present in combined data sets using cases where measurements for the same compound are reported in multiple assays against the same target. This approach shows that IC50assays selected using minimal curation settings have poor agreement with each other: almost 65% of the points differ by more than 0.3 log units, 27% differ by more than one log unit, and the correlation between the assays, as measured by Kendall’s τ, is only 0.51. Requiring that most of the assay metadata in ChEMBL matches (“maximal curation”) in order to combine two assays improves the situation (48% of the points differ by more than 0.3 log units, 13% by more than one log unit, and Kendall’s τ is 0.71) at the expense of having smaller data sets. Surprisingly, our analysis shows similar amounts of noise when combining data from different literature Kiassays. We suggest that good scientific practice requires careful curation when combining data sets from different assays and hope that our maximal curation strategy will help to improve the quality of the data that are being used to build and validate ML models for bioactivity prediction. To help achieve this, the code and ChEMBL queries that we used for the maximal curation approach are available as open-source software in our GitHub repository, https://github.com/rinikerlab/overlapping_assays.