MMMathijs F. J. Mabesoone

Mathijs F. J. Mabesoone

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Mathijs F. J. Mabesoone has a total of 123 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    BigMixSolDB: extraction of a solubility database in solvent mixtures with an uncertainty-quantified large language model-based pipeline

    by Andrei Voinea, Anna C. M. Thöni, Elija Veenman, Wilhelm T. S. Huck, Tal Kachman, Mathijs F. J. Mabesoone

    The rapidly expanding body of chemical literature contains physical data scattered across unstructured text and complex tables in thousands of publications.

  • Open Access

    Solving the Computational Peptide Design Puzzle: Challenges and Opportunities

    by Daniela Kalafatovic, Aysenur Iscen, Mathijs F. J. Mabesoone, Ioana M. Ilie, Goran Mauša, Ivan R. Sasselli

    Peptides are increasingly recognized as promising therapeutics, antimicrobial agents, and supramolecular biomaterials due to the remarkable versatility encoded in their amino acid sequences. However, the rational design of peptides remains challenging because sequence-structure–function relationships are not yet fully understood. Experimental exploration of peptide chemical space is inherently limited by the vast number of possible sequences and the complexity of resolving peptide behavior at molecular resolution. Computational methods, particularly molecular dynamics simulations and, more recently, machine learning and deep learning, have emerged as promising strategies to address these limitations. Although computational objectives differ, ranging from binding poses and biological activity to self-assembly propensity and supramolecular morphologies, they share common challenges associated with methodological heterogeneity, incomplete experimental metadata, insufficient benchmarking, and the lack of standardized and interoperable data sets. Here, we present our perspective on the main challenges in the computational study and design of peptides and discuss how collective efforts toward standardized data generation and data-driven discovery frameworks may help advance the field.

  • Open Access

    Data-driven discovery of pH-dependent surfactant behavior in a peptide amphiphile library

    by Justin Houx, Marco J.W. Kong, Siyu Tao, Pim F.J. Dankloff, Gijs Martens, Joost Hoeksel, Daniel Gironés, William E. Robinson, Wilhelm T.S. Huck, Mathijs F.J. Mabesoone

    Developing bio-based formulations is essential for reducing the environmental impact of modern surfactants. Peptide amphiphile (PA) surfactants offer a modular platform for developing materials with tailored structures and functions, making them highly attractive to replace existing petrochemical surfactants. However, rationally designing peptide sequences with predictable and tunable surface activity remains challenging. Existing sequence-based models primarily target aggregation and antimicrobial activity and cannot predict surface activity or account for environmental conditions such as pH and temperature. To address this shortcoming, we combine experimental and computational approaches to investigate sequence–property relationships of PAs. We employ automated experimental workflows to generate a consistent dataset of PA surface tension across a range of pHs. Our analyses reveal that the presence of proline most strongly induces surface activity at all pHs. Positively charged residues such as lysine and arginine more strongly introduce surface activity than negatively charged residues. Experimental validation of a machine learning classifier trained on the data collected for our library showed that the machine learning model correctly predicts surface activity in 97% of the sampled predicted surface active sequences. Thereby, we show that machine learning models trained on our dataset generalize across the sequence space, and enable targeted design of surface active PAs in diverse environments.

  • Open Access

    BigMixSolDB: Extraction of a solubility database in solvent mixtures with an uncertainty-quantified large language model-based pipeline

    by Andrei Voinea, Anna C.M. Thöni, Elija Veenman, Wilhelm T.S. Huck, Tal Kachman, Mathijs F.J. Mabesoone

    The rapidly expanding body of chemical literature contains physical data scattered across unstructured text and complex tables in thousands of publications. Translating this information into machine-readable formats is essential for training data-driven chemical models. The lack of uncertainty quantification in computationally extracted datasets makes manual extraction and curation still necessary. We here address this challenge and present an uncertainty-quantified data extraction pipeline. Our fully automated pipeline utilizes large language models (LLMs) to extract complex tabular and textual chemical data directly from scientific PDFs. We systematically assess several modern LLMs and document-to-text conversion frameworks. We benchmark the uncertainty in extracted data by comparison of our computationally extracted databases with large, existing databases on solubility that are manually curated. Our pipeline achieves between 93 and 79% of the literature contained data, with near-zero median log deviations for structurally matched rows and residual high-magnitude errors concentrated in a small subset of OCR, table-structure, unit-scaling, or reference-dataset discrepancies. We showcase the utility of our pipeline by deploying it to extract data from over 2,000 articles and generate BigMixSolDB, a comprehensive database on solubility in complex mixtures. BigMixSolDB comprises 280,273 solubility entries spanning single, binary, and ternary solvent systems. Our results demonstrate how integrated LLM-based pipelines can be used for literature data extraction that is comparable with manual curation. We envision that these frameworks can be used for large-scale, accurate extraction of literature data into datasets for machine learning applications

  • Open Access

    BigMixSolDB: Extraction of a solubility database in solvent mixtures with an uncertainty-quantified large language model-based pipeline

    by Andrei Voinea, Anna C.M. Thöni, Elija Veenman, Wilhelm T.S. Huck, Tal Kachman, Mathijs F.J. Mabesoone

    The rapidly expanding body of chemical literature contains physical data scattered across unstructured text and complex tables in thousands of publications. Translating this information into machine-readable formats is essential for training data-driven chemical models. The lack of uncertainty quantification in computationally extracted datasets makes manual extraction and curation still necessary. We here address this challenge and present an uncertainty-quantified data extraction pipeline. Our fully automated pipeline utilizes large language models (LLMs) to extract complex tabular and textual chemical data directly from scientific PDFs. We systematically assess several modern LLMs and document-to-text conversion frameworks. We benchmark the uncertainty in extracted data by comparison of our computationally extracted databases with large, existing databases on solubility that are manually curated. Our pipeline achieves between 93 and 79% of the literature contained data, with near-zero median log deviations for structurally matched rows and residual high-magnitude errors concentrated in a small subset of OCR, table-structure, unit-scaling, or reference-dataset discrepancies. We showcase the utility of our pipeline by deploying it to extract data from over 2,000 articles and generate BigMixSolDB, a comprehensive database on solubility in complex mixtures. BigMixSolDB comprises 280,273 solubility entries spanning single, binary, and ternary solvent systems. Our results demonstrate how integrated LLM-based pipelines can be used for literature data extraction that is comparable with manual curation. We envision that these frameworks can be used for large-scale, accurate extraction of literature data into datasets for machine learning applications


  • Open Access

    BigMixSolDB: Extraction of a solubility database in solvent mixtures with an uncertainty-quantified large language model-based pipeline

    by Andrei Voinea, Anna C.M. Thöni, Elija Veenman, Wilhelm T.S. Huck, Tal Kachman, Mathijs F.J. Mabesoone

    The rapidly expanding body of chemical literature contains physical data scattered across unstructured text and complex tables in thousands of publications. Translating this information into machine-readable formats is essential for training data-driven chemical models. The lack of uncertainty quantification in computationally extracted datasets makes manual extraction and curation still necessary. We here address this challenge and present an uncertainty-quantified data extraction pipeline. Our fully automated pipeline utilizes large language models (LLMs) to extract complex tabular and textual chemical data directly from scientific PDFs. We systematically assess several modern LLMs and document-to-text conversion frameworks. We benchmark the uncertainty in extracted data by comparison of our computationally extracted databases with large, existing databases on solubility that are manually curated. Our pipeline achieves between 93 and 79% of the literature contained data, with extraction errors being on par with aleatoric limits that exist in the manually curated databases. We showcase the utility of our pipeline by deploying the optimized pipeline to extract data from 2,301 articles to generate BigMixSolDB, a comprehensive database on solubility in complex mixtures. BigMixSolDBcomprises over 299,000 filtered solubility entries spanning single, binary, and ternary solvent systems. Our results demonstrate how integrated LLM-based pipelines can be used for literature data extraction that is comparable with manual curation. We envision that these frameworks can be used for large-scale, accurate extraction of literature data into datasets for machine learning applications.

  • Solvent-Induced Stereomutation in Supramolecular Assemblies Explained by Hansen Solubility Parameters

    by Magda M. J. Dekker, Dibyojeet Bagchi, Rense Terpstra, Maarten Bransen, Mathijs F. J. Mabesoone, Heiner Friedrich, Ghislaine Vantomme, E. W. Meijer

    Solvent-induced stereomutation, the reversal of supramolecular helicity driven by solvent identity, is a striking yet poorly understood phenomenon in supramolecular chemistry, with implications for chiral materials, molecular recognition, and functional assemblies. Despite numerous observations in diverse systems, a general framework for understanding and controlling such helicity inversion between supramolecular diastereomers remains elusive. Here, we present a quantitative strategy based on Hansen solubility parameters (HSPs) that identifies distinct and reproducible solvent regimes associated with opposite helicities in two structurally different chiral monomers: a C2-symmetric perylene diimide derivative (S -PDI) with chiral citronellol side chains attached through gallate wedges, and a C3-symmetric truxenone derivative (S -TTA) bearing the same chiral side chains connected via hydrogen-bonding amides. By mapping supramolecular handedness within the HSP space, we show that helicity arises not only from molecular structure but also from the interplay between solvent dispersion, polarity, and hydrogen-bonding, which modulate the solvation of side chains and influence the geometry of aromatic stacking. Furthermore, the helicity induced by the solvent environment is not a result of variations in the assembly pathway and is independent of concentration. These results establish HSP mapping as a general strategy for explaining the opposite supramolecular chirality controlled by the solvent and for designing adaptive, reconfigurable, and functional chiral materials.


  • Open Access

    BigMixSolDB: Extraction of a solubility database in solvent mixtures with an uncertainty-quantified large language model-based pipeline

    by Andrei Voinea, Anna C.M. Thöni, Elija Veenman, Wilhelm T.S. Huck, Tal Kachman, Mathijs F.J. Mabesoone

    The rapidly expanding body of chemical literature contains physical data scattered across unstructured text and complex tables in thousands of publications. Translating this information into machinereadable formats is essential for training data-driven chemical models. The lack of uncertainty quantification in computationally extracted datasets make manual extraction and curation still necessary. We here address this challenge and present an uncertainty-quantified data extraction pipeline. Our fully automated pipeline utilizes large language models (LLMs) to extract complex tabular and textual chemical data directly from scientific PDFs. We systematically assess several modern LLMs and document-to-text conversion frameworks. We benchmark the uncertainty in extracted data by comparison of our computationally extracted databases with large, existing databases on solubility that are manually curated. Our pipeline achieve between 93 and 79% of the literature contained data, with extraction errors being on par with aleatoric limits that exist in the manually curated databases. We showcase the utility of our pipeline by deploying the optimized pipeline across a diverse corpus of 2,793 articles to generate BigMixSolDB, a comprehensive database on solubility in complex mixtures. BigMixSolDB comprises over 325,000 filtered solubility entries spanning single, binary, and ternary solvent systems. Our results demonstrate how integrated LLM-based pipelines can be used for literature data extraction that is comparable with manual curation. We envision these frameworks can be used for large-scale, accurate extraction of literature data into datasets for machine learning applications.

  • Top Cited Papers

    Evolution-guided engineering of trans -acyltransferase polyketide synthases

    by Mathijs F. J. Mabesoone, Stefan Leopold-Messer, Hannah A. Minas, Clara Chepkirui, Pornsuda Chawengrum, Silke Reiter, Roy A. Meoded, Sarah Wolf, Ferdinand Genz, Nancy Magnus, et al.

    Bacterial multimodular polyketide synthases (PKSs) are giant enzymes that generate a wide range of therapeutically important but synthetically challenging natural products. Diversification of polyketide structures can be achieved by engineering these enzymes. However, notwithstanding successes made with textbook cis -acyltransferase ( cis -AT) PKSs, tailoring such large assembly lines remains challenging. Unlike textbook PKSs, trans -AT PKSs feature an extraordinary diversity of PKS modules and commonly evolve to form hybrid PKSs. In this study, we analyzed amino acid coevolution to identify a common module site that yields functional PKSs. We used this site to insert and delete diverse PKS parts and create 22 engineered trans -AT PKSs from various pathways and in two bacterial producers. The high success rates of our engineering approach highlight the broader applicability to generate complex designer polyketides.


  • Structure of a Promiscuous Thioesterase Domain Responsible for Branching Acylation in Polyketide Biosynthesis

    by Amy E. Fraley, Cora L. Dieterich, Mathijs F. J. Mabesoone, Hannah A. Minas, Roy A Meoded, Franziska Hemmerling, Jörn Piel

    Thioesterases (TEs) are fundamentally important enzymes present in all bacteria and eukaryotes, where they have conserved functions in fatty acid biosynthesis and secondary metabolism. This work provides the first structural insights into a functionally distinct group of TEs that perform diverse acylations in polyketide and peptide biosynthesis (TEBs). Structural analysis of the oocydin (OocS) TEB domain facilitated identification and engineering of the active site to modulate acyl‐group acceptance. In this way, we achieved higher reactivity using a structure‐based approach, building a foundation for biocatalytic development of TEB‐mediated O‐acylation, a modification known to improve the bioactivity of oocydin‐type polyketides. Lastly, the promiscuity of the OocS TEB motivated us to investigate, and ultimately provide evidence for, the production of longer chain branched oocydins in the native host Serratia plymuthica 4Rx13. This work frames the OocS TEB and homologs as invaluable synthetic biology tools for polyketide drug development.