CSCarl C. G. Schiebroek

Carl C. G. Schiebroek

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Carl C. G. Schiebroek has a total of 27 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Customizable Approach to Conformer Space Discretization using Torsion Angular Bin Strings (customTABS)

    by Jessica Braun, Enrico Ruijsenaars, Carl Schiebroek, Djahan Lamei, Paul Katzberger, Gregory A. Landrum, Sereina Riniker

    Conformational ensembles of small organic molecules are primarily determined by the rotameric states of their torsional or dihedral angles. The previously introduced concept of torsion angular bin strings (TABS) is a discrete vector representation of a molecule’s dihedral states, which therefore offer a quantitative description of such ensembles. Discretization of the torsional-angle space allows us to count distinct conformational states, giving an estimate of the upper limit of the expected conformational ensemble size (nTABS). In addition to nTABS being used as a qualitative measure of molecular flexibility, TABS itself is a way of grouping conformers of a molecule without picking thresholds. In the initial version of the TABS application, the discretization was based on crystal-structure derived torsion preferences, which may not be fully transferable to the conformational preferences in other environments. Here, we present the extension of the TABS framework to consider observed, system-specific distributions for the discretization of the conformational space. CustomTABS can be used to decompose the given conformational ensemble into the torsional states, which can be useful for clustering, importance ranking of dihedrals, or convergence analysis. We present the algorithmic solutions developed for the customTABS extension and demonstrate their application in clustering molecular dynamics (MD) trajectories and monitoring the convergence of the observed torsional space.

  • Open Access

    Balancing Data Quantity and Quality: Evaluating Curation Strategies for Bioactivity Prediction in Lead Optimization

    by Carl C. G. Schiebroek, Gregory A. Landrum, Sereina Riniker

    Building good machine-learning (ML) models to predict the bioactivity of novel chemical matter remains a challenging task. Accurate models require a training set with a large number of diverse compounds and a low level of noise. When extracting data from public databases such as ChEMBL, different levels of curation rigor may be applied, resulting in training sets of varying size, diversity, and, presumably, noise levels. It is not possible to know a priori whether increasing the size of the data set at the cost of adding more noise improves model generalization. To assess this trade-off, we compare three data curation and modeling approaches: (1) models trained on data for a single target, (2) models trained on target-specific data further restricted to a single set of assay conditions, and (3) multitask learning (MTL) models where each assay condition is treated as a separate task. This MTL approach was designed to bridge the gap between data quantity and quality. Graph neural networks (GNN) and random forests (RF) regressors are evaluated via a leave-assay-out strategy to minimize noise in the test sets. Our results show no meaningful performance differences between these curation strategies, suggesting that for lead-optimization tasks, increasing data quantity at the expense of label consistency does not improve generalization. Notably, the MTL approach also failed to provide a performance advantage. Additionally, we find that GNNs exhibit high seed-dependent variability in connection with the comparatively small training sets common for bioactivity measurements, highlighting the necessity of multiseed evaluation for robust model assessment.


  • Open Access

    Balancing Data Quantity and Quality: Evaluating Curation Strategies for Bioactivity Prediction

    by Carl C. G. Schiebroek, Gregory A. Landrum, Sereina Riniker

    Building good machine-learning (ML) models to predict the bioactivity of novel chemical matter remains a challenging task. Accurate models require a training set with a large number of diverse compounds and a low level of noise. When extracting data from public databases such as ChEMBL, different levels of curation rigor may be applied, resulting in training sets of varying size, diversity, and, presumably, noise levels. It is not possible to know a priori whether increasing the size of the dataset at the cost of adding more noise improves model generalization. To assess this trade-off, we compare three data curation and modeling approaches: (1) models trained on data for a single target, (2) models trained on target-specific data further restricted to a single set of assay conditions, and (3) multitask learning (MTL) models where each assay condition is treated as a separate task. This MTL approach was designed to bridge the gap between data quantity and quality. Graph neural networks (GNN) and random forests (RF) regressors are evaluated via a leave-assay-out strategy to minimize noise in the test sets. Our results show no meaningful performance differences between these curation strategies, suggesting that for lead-optimization tasks, increasing data quantity at the expense of label consistency does not improve generalization. Notably, the MTL approach also failed to provide a performance advantage. Additionally, we find that GNNs exhibit high seed-dependent variability in connection with the comparatively small training sets common for bioactivity measurements, highlighting the necessity of multi-seed evaluation for robust model assessment.

  • Open Access

    Author Correction: Prospective de novo drug design with deep interactome learning

    by Kenneth Atz, Leandro Cotos, Clemens Isert, Maria Håkansson, Dorota Focht, Mattis Hilleke, David F. Nippa, Michael Iff, Jann Ledergerber, Carl C. G. Schiebroek, et al.


  • Open Access

    Protein Binding Site Representation in Latent Space

    by Frederieke Lohmann, Stephan Allenspach, Kenneth Atz, Carl C. G. Schiebroek, Jan A. Hiss, Gisbert Schneider

    Interpretability and reliability of deep learning models are important for computer‐based drug discovery. Aiming to understand feature perception by such a model, we investigate a graph neural network for affinity prediction of protein‐ligand complexes. We assess a latent representation of ligand binding sites and investigate underlying geometric structure in this latent space and its relation to protein function. We introduce an automated computational pipeline for dimensionality reduction, clustering, hypothesis testing, and visualization of latent space. The results indicate that the learned protein latent space is inherently structured and not randomly distributed. Several of the identified protein binding site clusters in latent space correspond to functional protein families. Ligand size was found to be a determinant of cluster geometry. The computational pipeline proved applicable to latent space analysis and interpretation and can be adapted to work for different datasets and deep learning models.


  • Top Cited Papers
    Open Access

    Prospective de novo drug design with deep interactome learning

    by Kenneth Atz, Leandro Cotos, Clemens Isert, Maria Håkansson, Dorota Focht, Mattis Hilleke, David F. Nippa, Michael Iff, Jann Ledergerber, Carl C. G. Schiebroek, et al.

    De novo drug design aims to generate molecules from scratch that possess specific chemical and pharmacological properties. We present a computational approach utilizing interactome-based deep learning for ligand- and structure-based generation of drug-like molecules. This method capitalizes on the unique strengths of both graph neural networks and chemical language models, offering an alternative to the need for application-specific reinforcement, transfer, or few-shot learning. It enables the “zero-shot" construction of compound libraries tailored to possess specific bioactivity, synthesizability, and structural novelty. In order to proactively evaluate the deep interactome learning framework for protein structure-based drug design, potential new ligands targeting the binding site of the human peroxisome proliferator-activated receptor (PPAR) subtype gamma are generated. The top-ranking designs are chemically synthesized and computationally, biophysically, and biochemically characterized. Potent PPAR partial agonists are identified, demonstrating favorable activity and the desired selectivity profiles for both nuclear receptors and off-target interactions. Crystal structure determination of the ligand-receptor complex confirms the anticipated binding mode. This successful outcome positively advocates interactome-based de novo design for application in bioorganic and medicinal chemistry, enabling the creation of innovative bioactive molecules.


  • Open Access

    DASH: Dynamic Attention-Based Substructure Hierarchy for Partial Charge Assignment

    by Marc T. Lehner, Paul Katzberger, Niels Maeder, Carl C.G. Schiebroek, Jakob Teetz, Gregory A. Landrum, Sereina Riniker

    We present a robust and computationally efficient approach for assigning partial charges of atoms in molecules. The method is based on a hierarchical tree constructed from attention values extracted from a graph neural network (GNN), which was trained to predict atomic partial charges from accurate quantum-mechanical (QM) calculations. The resulting dynamic attention-based substructure hierarchy (DASH) approach provides fast assignment of partial charges with the same accuracy as the GNN itself, is software-independent, and can easily be integrated in existing parametrization pipelines, as shown for the Open force field (OpenFF). The implementation of the DASH workflow, the final DASH tree, and the training set are available as open source/open data from public repositories.