Igor Kozlovskii
  • Home
  • Research Blog
  • Notes
  • CV

On this page

  • Curriculum Vitae
    • Education
    • Experience
    • Selected Publications
    • Research Projects
    • Technical Skills
    • Teaching
    • Scholarships & Awards
    • Patents and Software

CV

Curriculum Vitae

Last updated: November 2025

A full PDF CV is available on request.

Education

  • PhD in Computer Science, Constructor University Bremen, Germany
    2024 – present
    • Research group of Prof. Petr Popov
    • Focus: structural bioinformatics, deep learning, generative models for drug design.
  • PhD (Candidate of Sciences), Skolkovo Institute of Science and Technology (Skoltech), Russia
    2020 – 2024
    • Computational and Data Science and Engineering (CDSE), iMolecule lab
    • Thesis: deep learning for binding site identification in macromolecules.
  • MSc in Applied Physics and Mathematics, Moscow Institute of Physics and Technology (MIPT), Russia
    2018 – 2020
    • Biophysics track, Laboratory of Structural Biology of G Protein-Coupled Receptors
    • GPA: 4.83 / 5.
  • BSc in Applied Physics and Mathematics, Moscow Institute of Physics and Technology (MIPT), Russia
    2014 – 2018
    • Biophysics track, Department of General and Applied Physics
    • GPA: 4.38 / 5.

Experience

  • Researcher / Software Developer, Tetra D AG, Switzerland
    2024 – present
    • Development of a computational platform for structure-based drug design.
    • Design and implementation of hierarchical 3D latent diffusion models for molecule generation.
  • Researcher / PhD Student, iMolecule lab, Skoltech
    2019 – 2024
    • Deep learning for binding site identification in proteins, nucleic acids, and complexes.
    • Large-scale dataset curation, model development (BiteNet family), and case studies on diverse targets.
  • Researcher (BSc/MSc Student), Laboratory of Structural Biology of GPCRs, MIPT
    2017 – 2020
    • Binding-site detection on protein complexes using 3D CNNs.
    • Identification of stabilizing mutations in GPCRs using machine learning.

Selected Publications

  • Computational methods for binding site prediction on macromolecules
    Kozlovskii, I., Popov, P.
    Quarterly Reviews of Biophysics, 58, e12, 2025.
    10.1017/S003358352500006X

  • Approaching Optimal pH Enzyme Prediction with Large Language Models
    Zaretckii, M., Buslaev, P., Kozlovskii, I., Morozov, A., Popov, P.
    ACS Synthetic Biology, 13(9), 3013–3021, 2024.
    10.1021/acssynbio.4c00465

  • Unraveling viral drug targets: a deep learning-based approach for the identification of potential binding sites
    Popov, P., Kalinin, R., Buslaev, P., Kozlovskii, I., Zaretckii, M., Karlov, D., … Stepanov, A.
    Briefings in Bioinformatics, 25(1), bbad459, 2024.
    10.1093/bib/bbad459

  • A community effort in SARS-CoV-2 drug discovery
    Schimunek, J., Seidl, P., Elez, K., Hempel, T., Le, T., Noé, F., … Kozlovskii, I., … Varnek, A.
    Molecular Informatics, 43(1), e202300262, 2024.
    10.1002/minf.202300262

  • Structure-based deep learning for binding site detection in nucleic acid macromolecules
    Kozlovskii, I., & Popov, P.
    NAR Genomics and Bioinformatics, 3(4), lqab111, 2021.
    10.1093/nargab/lqab111

  • Protein–peptide binding site detection using 3D convolutional neural networks
    Kozlovskii, I., & Popov, P.
    Journal of Chemical Information and Modeling, 61(8), 3814–3823, 2021.
    10.1021/acs.jcim.1c00475

  • Spatiotemporal identification of druggable binding sites using deep learning
    Kozlovskii, I., & Popov, P.
    Communications Biology, 3, 618, 2020.
    10.1038/s42003-020-01350-0

Research Projects

  • Binding site detection with 3D CNNs (BiteNet family)
    • Fast voxelization of protein atomic environments (C++/CUDA, OpenMP).
    • 3D YOLO-like architectures for binding site detection on proteins, peptides, and nucleic acids.
    • Large-scale PDB-derived datasets with sequence/structure/site-based de-duplication.
  • Hierarchical 3D latent diffusion for molecule design
    • Hierarchical pipeline combining VAE, diffusion model, and reconstruction modules for 3D molecules.
    • Focus on structure-based generative design for drug discovery.
  • ML for thermostabilizing mutations in GPCRs
    • Structural and physicochemical descriptor pipelines for mutated positions.
    • Ensemble models for classifying stabilizing vs non-stabilizing mutations.

Technical Skills

  • Programming: Python, C++, CUDA
  • Deep learning: PyTorch, TensorFlow, PyTorch Lightning
  • Scientific stack: NumPy, SciPy, pandas, scikit-learn, matplotlib, seaborn, numba, polars, plotly
  • Bioinformatics / cheminformatics: Biopython, MDAnalysis, RDKit, Modeller, PyRosetta, PDBFixer, parasail
  • Simulation & modeling tools: GROMACS, ICM-Pro, HH-suite, MMseqs2, FoldSeek, MAFFT, BLAST, Rosetta, AlphaFold2
  • Molecular visualization: PyMOL, VMD, SAMSON, py3Dmol
  • Tooling & infrastructure: Git, Docker, Slurm, Conda, pyenv, Poetry, Pixi, Flask

Teaching

  • Teaching Assistant, “Business Communication” course, Skoltech (2023)
  • Guest lectures on machine learning topics (decision trees, ensembles, etc.) at university level.

Scholarships & Awards

  • Presidential Scholarship of the Russian Federation, 2022–2023
  • Copenhagen Bioinformatics Hackathon 2021 – 1st place (optimal pH prediction challenge).
  • Health Data Hack 2022 – 3rd place (cancer cell detection in histopathology, CV track).
  • Other hackathons and competitions are listed in the full CV.

Patents and Software

  • Certificate of State Registration of a Computer Program No. RU 2023613561 – BiteNet (v1.2), 2023.
  • Patent No. RU 2743316 – Method for identification of binding sites of protein complexes, 2021.
  • Certificate of State Registration of a Computer Program No. RU 2020661135 – BiteNet, 2020.