Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Pavel Polishchuk has a total of 207 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Synthesis and Biological Activity of 2,6‐Disubstituted and 2,6,9‐Trisubstituted 7‐Deazapurine Nucleobases

    by Ugnė Šinkevičiūtė, KateřIna Burgetová Ječmeňová, Soňa Gurská, Lenka Poštová Slavětínská, Vincent Baran, Pavel Polishchuk, Petr Džubák, Michal Hocek, Marián Hajdúch, Michal Tichý

    High‐throughput screening of the IMTM compound library for modulatory activity on adenosine receptors (ARs) identified two deazapurine hits based on 2,6‐diaryl‐7‐deazapurines. In this study, a series of 16 new 2,6‐diaryl‐7‐deazapurine nucleobases together with four N‐9 substituted derivatives were synthesized. The synthesis was based on regioselective Suzuki cross‐coupling into position 6 followed by a second Suzuki cross‐coupling at position 2 under harsher conditions. Both cross‐couplings were performed on N‐9 protected nucleobases. Biological activity profiling identified several potent antagonists of all four subtypes of ARs and four compounds as sub‐micromolar and very selective antagonists of A 3 . Several final compounds showed potent in vitro cytotoxic activity on a small panel of cancer cell lines, but with no selectivity against nonmalignant fibroblasts BJ and MRC‐5.

  • Open Access

    CReM update: a faster, more compact, and more flexible framework for chemically reasonable mutations

    by Pavel Polishchuk

    CReM is a context-based molecular structure generation framework that produces chemically reasonable structures by exchanging fragments observed in equivalent local environments. Here, we present an updated CReM version that improves scalability, speed, and flexibility while preserving backward compatibility. The new implementation introduces intramolecular cyclization and partial ring replacement, going beyond classical matched molecular pair replacements and enabling more flexible exploration of cyclic and macrocyclic chemical space. A normalized database schema reduces storage redundancy, supports multiple named fragment sets in a single database, and enables property-aware fragment filtering. Database generation is accelerated through parallel processing and merging, and a Python API was implemented for programmatic workflows. Replacement of the previous reaction-based assembly backend with an RDKit molzip-based procedure further improves structure generation speed. Benchmarks using ChEMBL36 and SureChEMBL fragment databases demonstrate reduced database size, faster database construction and structure generation, and controllable synthetic accessibility through context radius and fragment database composition. Scientific contribution This work extends the CReM structure generation framework by addressing several limitations of the original approach, particularly in the generation of ring systems. The proposed expansion improves both database construction and structure generation speed, while also optimizing database storage. These advances enable the use of substantially larger fragment databases, thereby allowing broader coverage of chemical space than was previously feasible. In addition, the framework provides greater flexibility in molecular structure generation while preserving indirect control over the synthetic accessibility of the generated compounds.

  • Open Access

    EasyDock 1.3: An Automated Pipeline for Molecular Docking

    by Guzel Minibaeva, Veincent Yap, Pavel Polishchuk

    Molecular docking is widely used in drug design, particularly for large compound libraries. We previously developed EasyDock, an automated docking pipeline with multiserver task distribution to support such large-scale campaigns, and present here its extended version. Supported docking engines now include Vina-family CPU- and GPU-based variants (QVina2, Vina-GPU, etc.) and deep-learning-based engines (CarsiDock, SurfDock) via a built-in client–server architecture. Ligand preparation was enriched with salt stripping, stereoisomer enumeration, and conformational sampling of saturated ring systems. Integration of open-source protonation tools (pkasolver, MolGpKa, Uni-pKa) replaces previously required commercial software, making the pipeline fully open source. Postdocking analysis now includes protein–ligand interaction fingerprint (PLIF) computation and pose quality assessment via PoseBusters. We provide Apptainer/Docker containers for the docking engines and protonation tools, simplifying installation and HPC deployment. The source code is available at https://github.com/ci-lab-cz/easydock.


  • Open Access

    CReM-dock: de novo design of synthetically feasible structures guided by molecular docking

    by Guzel Minibaeva, Haolin Du, Finlay Clark, Julien Michel, Pavel Polishchuk

    The de novo generation of chemical compounds represents a compelling strategy for the exploration of a significantly broader chemical space compared to traditional virtual screening methods.


  • Open Access

    CReM-pharm: de novo 3D pharmacophore-based design with synthetic accessibility awareness

    by Alina Denzler, Dinesh Kumar Sriramulu, Jozef Pecha, Pavel Polishchuk

    De novo design methodologies have the potential to significantly enhance the exploration of chemical space in the search for promising ligands featuring no


  • Open Access

    Thiazolidinone-based compounds as dual-purpose therapeutics: antimicrobial efficacy, cytotoxicity and pharmacokinetic potential

    by Dmytro Mural, Dmytro Khyluk, Andrii Lozynskyi, Victoriya Georgiyants, Olexandra Roman, Anna Kryshchyshyn-Dylevych, Sona Gurska, Pavel Polishchuk, Petr Dzubak, Marian Hajduch, et al.

    Infectious diseases and cancer remain leading global health challenges, with rising resistance to existing antibiotics and limited selectivity of many cytotoxic agents. Heterocyclic scaffolds, particularly thiazolidinones, offer a promising platform for the development of novel antimicrobial and anticancer compounds. The aim of the study. To evaluate the antimicrobial and cytotoxic properties of thiazolidinone-based compounds against a panel of pathogenic microorganisms and human cancer cell lines, and to identify the most promising derivatives with favorable safety, pharmacokinetic, and mechanistic profiles through molecular docking and dynamics studies. Materials and methods. A library of 5-enamine(hydrazine)-4-thiazolidinone derivatives was screened for antimicrobial activity against Gram-positive and Gram-negative bacteria and Candida albicans, and for cytotoxic activity against six human cancer cell lines. Minimum inhibitory concentrations (MIC) were determined, and IC₅₀ values were measured for selected compounds. Pharmacokinetic properties, including gastrointestinal absorption and lipophilicity, were assessed in silico. To investigate potential mechanisms of antibacterial action, molecular docking was performed against MurB (UDP-N-acetylenolpyruvylglucosamine reductase) and DNA gyrase subunit B (ATPase domain), followed by molecular dynamics (MD) simulations to evaluate the stability of the most promising complexes. Results. Thirty-two compounds exhibited antimicrobial activity (MIC ≤ 200 µM), and ten (6, 7, 10, 12, 13, 16, 19, 21, 22, and 29) were identified as the most active. Compound 29, an isatin–oxadiazole hybrid, demonstrated potent activity against Enterococcus faecalis and vancomycin-resistant E. faecium (MIC = 3.13 µM), outperforming vancomycin. Compound 21 was highly active against Staphylococcus epidermidis (MIC = 1.56 µM), while compound 6 showed efficacy against methicillin-susceptible and -resistant S. aureus (MIC = 6.25 µM). Moderate antifungal activity was observed for compound 27 (MIC = 100 µM), whereas Gram-negative bacteria were largely resistant. Cytotoxicity screening revealed selective anticancer activity of compounds 12 and 27, with high therapeutic indices against CCRF-CEM cells and minimal effects on normal fibroblasts. Compound 2 exhibited strong cytotoxicity (IC₅₀ = 1.1 µM), while compound 29 combined non-cytotoxicity with favorable pharmacokinetic characteristics. Molecular docking supported MurB as the primary antibacterial target, with the most active compounds (21 and 29) showing the most favorable binding energies. Compound 29 also exhibited strong affinity for GyrB, indicating a potential dual-target mechanism. Molecular dynamics confirmed that MurB–compound 29 complexes were particularly stable, correlating well with experimental antibacterial activity. Conclusions. Thiazolidinone-based hybrids demonstrated promising antimicrobial and anticancer properties. Compound 29 emerged as a particularly attractive dual-purpose candidate due to its potent activity, safety profile, favorable pharmacokinetics, and validated interaction with essential bacterial enzymes. Together, biological and computational results support the potential of thiazolidinone scaffolds as a basis for the development of selective or multitarget therapeutic agents

  • Open Access

    Triterpenoid derivatives inhibit Gli-mediated transcription in human glioblastoma cell line via direct interaction with Gli1

    by Ivo Frydrych, Jiří Řehulka, Lucie Slavíková, Jiří Hodoň, Jan Pokorný, Veronika Šidová, Martina Medvedíková, Hanuš Slavík, Pavel Polishchuk, Tomáš Oždian, et al.

  • Open Access

    CReM-dock: de novo design of synthetically feasible compounds guided by molecular docking

    by Guzel Minibaeva, Pavel Polishchuk

    De novo generation of compounds is an attractive strategy allowing to explore much broader chemical space than virtual screening. Fragment-based approaches suffer from low synthetic accessibility of generated compounds. In this study we combined the previously developed fragment-based generator CReM and molecular docking to guide the exploration of chemical space. The developed approach is very flexible and allows to indirectly control synthetic accessibility of generated compounds, their diversity by choosing one of selection strategies, augmentation of an objective function to generate more drug-like compounds, control over preserving important protein-ligand interactions and ligand poses. The generated compounds demonstrated high novelty and were competitive to compounds generated with REINVENT4. We demonstrated in different case studies flexibility of the developed approach and its applicability to de novo generation as well as fragment expansion tasks.


  • Open Access

    Correction: StreaMD: the toolkit for high-throughput molecular dynamics simulations

    by Aleksandra Ivanova, Olena Mokshyna, Pavel Polishchuk


  • Open Access

    StreaMD: the toolkit for high-throughput molecular dynamics simulations

    by Aleksandra Ivanova, Olena Mokshyna, Pavel Polishchuk

    Molecular dynamics simulations serve as a prevalent approach for investigating the dynamic behaviour of proteins and protein–ligand complexes. Due to its versatility and speed, GROMACS stands out as a commonly utilized software platform for executing molecular dynamics simulations. However, its effective utilization requires substantial expertise in configuring, executing, and interpreting molecular dynamics trajectories. Existing automation tools are constrained in their capability to conduct simulations for large sets of compounds with minimal user intervention, or in their ability to distribute simulations across multiple servers. To address these challenges, we developed a Python-based tool that streamlines all phases of molecular dynamics simulations, encompassing preparation, execution, and analysis. This tool minimizes the required knowledge for users engaging in molecular dynamics simulations and can efficiently operate across multiple servers within a network or a cluster. Notably, the tool not only automates trajectory simulation but also facilitates the computation of free binding energies for protein–ligand complexes and generates interaction fingerprints across the trajectory. Our study demonstrated the applicability of this tool on several benchmark datasets. Additionally, we provided recommendations for end-users to effectively utilize the tool.Scientific contributionThe developed tool, StreaMD, is applicable to different systems (proteins, ligands and their complexes including co-factors) and requires a little user knowledge to setup and run molecular dynamics simulations. Other features of StreaMD are seamless integration with calculation of MM-GBSA/PBSA binding free energies and protein-ligand interaction fingerprints, and running of simulations within distributed environments. All these will facilitate routine and massive molecular dynamics simulations.