MCMarcin Cieślak

Marcin Cieślak

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

Marcin Cieślak has a total of 6 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Qsarna: An Online Tool for Smart Chemical Space Navigation in Drug Design

    by Marcin Cieślak, Jan Łęski, Olga Krzysztyńska-Kuleta, Justyna Kalinowska-Tłuścik, Tomasz Danel

    Drug discovery is a lengthy and resource-intensive process that requires innovative computational techniques to expedite the transition from laboratory research to life-saving medications. Here, we introduce Qsarna, a comprehensive online platform that combines machine learning for activity prediction with traditional molecular docking to streamline virtual screening workflows. Our platform employs a fragment-based generative model, enabling the exploration of novel chemical spaces with the desired pharmacophoric features. Users can share results with others, and docking poses can be examined directly within the platform. In our case study, we successfully identified three new hits for monoamine oxidase B with nanomolar potency, which were later confirmed by experimental assays. The user-friendly web interface requires minimal computational expertise, making advanced virtual screening accessible to scientists regardless of their main field of study. Qsarna represents a significant advancement in computational drug discovery by seamlessly integrating complementary in silico approaches and democratizing access to advanced virtual screening technologies.


  • Open Access

    Structure-guided fragment linking algorithm enables chemically feasible ligand design with predicted binding modes

    by Marcin Cieślak, Tomasz Danel, Justyna Kalinowska-Tłuścik

    Fragment-based drug discovery offers an efficient strategy for exploring chemical space and identifying novel therapeutics. However, the rational design of linkers connecting promising fragments remains a significant bottleneck, with existing computational methods suffering from limited synthetic accessibility and inadequate preservation of fragment binding orientations. We developed CRET (Chemically Reasonable Tethering), a unified computational framework that automates the process of fragment linking by generating specific 3D conformations of compounds directly within the binding pocket. The algorithm identifies linkers that match the molecular context of both fragments and applies chemically plausible transformations to ensure high synthetic accessibility. The resulting molecule is then aligned with the fragment poses using template docking. Comprehensive benchmarking against established generative models demonstrates CRET's superior performance in the generation of unique, high-quality linkers, predicted binding pose accuracy, and preservation of fragment orientations. The developed algorithm was tested on challenging ligand classes such as PROTACs and macrocycles, yielding satisfactory results in terms of the generated structures within relevant protein binding sites. The strong results and the ability of the algorithm to handle complex ligands position CRET as a valuable tool for advancing structure-based drug discovery efforts.

  • Open Access

    Machine learning accelerates pharmacophore-based virtual screening of MAO inhibitors

    by Marcin Cieślak, Tomasz Danel, Olga Krzysztyńska-Kuleta, Justyna Kalinowska-Tłuścik

    Nowadays, an efficient and robust virtual screening procedure is crucial in the drug discovery process, especially when performed on large and chemically diverse databases. Virtual screening methods, like molecular docking and classic QSAR models, are limited in their ability to handle vast numbers of compounds and to learn from scarce data, respectively. In this study, we introduce a universal methodology that uses a machine learning-based approach to predict docking scores without the need for time-consuming molecular docking procedures. The developed protocol yielded 1000 times faster binding energy predictions than classical docking-based screening. The proposed predictive model learns from docking results, allowing users to choose their preferred docking software without relying on insufficient and incoherent experimental activity data. The methodology described employs multiple types of molecular fingerprints and descriptors to construct an ensemble model that further reduces prediction errors and is capable of delivering highly precise docking score values for monoamine oxidase ligands, enabling faster identification of promising compounds. An extensive pharmacophore-constrained screening of the ZINC database resulted in a selection of 24 compounds that were synthesized and evaluated for their biological activity. A preliminary screen discovered weak inhibitors of MAO-A with a percentage efficiency index close to a known drug at the lowest tested concentration. The approach presented here can be successfully applied to other biological targets as target-specific knowledge is not incorporated at the screening phase.