MSc student at Gdańsk University of Technology, working where computational chemistry meets bioinformatics.
I work on:
- cheminformatics and QSAR/QSPR modelling
- omics and chemometric data analysis
- small automation tools that save time
Toolbox:
- Languages: Python (pandas, NumPy, SciPy, scikit-learn, RDKit, matplotlib), SAS, MATLAB, SQL / PostgreSQL
- Chemistry and data tools: Orange Data Mining, PaDEL-Descriptor, MOPAC, CCDC Mercury, CrystalExplorer, PubChem / NCBI tools
| Project | Course | What's inside |
|---|---|---|
| qspr-microplastic-sorption | QSAR/QSPR Modelling | MLR, PLS, SVR, k-NN and decision trees predicting pollutant sorption on PP microplastics. The full 7-stage model-development history is included. Final MLR: Q²ext 0.91 ± 0.04 over 200 splits |
| lipidomics-human-milk-untargeted | Lipidomics and Glycomics | Untargeted LC-MS workflow: QC/RSD filtering (752 → 94 features), PCA, paired statistics with FDR correction, conference poster |
| cheminformatics-chemical-space | Cheminformatics | RDKit descriptors and fingerprints, PCA / t-SNE / UMAP, clustering, PubChem pipeline |
| biochemical-process-modeling | Modelling of Biochemical Processes | Kinetics, CSTR and Monod bioreactor fitting with bootstrap; spectral pKa and speciation |
| statistics-coursework | Biostatistics, Statistical Programming | SAS analysis (MTHFR genotype, homocysteine, lung cancer); GC-MS chromatogram processing in MATLAB |
| open-databases-structural-analysis | Open Databases | Crystal structure of a cisplatin analogue (CSD, Hirshfeld surfaces); PCR primer design for M. canis |
| postgresql-database-apps | Data Sets and Their Protection | PostgreSQL schema design and Tkinter + psycopg2 clients |
| delhi-climate-analysis | Python Language | pandas cleaning, aggregation and seasonality analysis |
- moodle-sync: Moodle materials, deadlines and grades synced to your computer, Google Drive and Calendar, with notifications.
- amp-cubecoders-tg-bot: Telegram bot for AMP game servers (player activity, stats, alerts, self-update).
- family-pet-bot: an AI "pet" for family chats on Telegram or Discord.
I use AI tools (large language models) as a coding assistant, and most of the code in these repositories was written with their help. My part is defining the problem, running and checking the analyses, interpreting the results and deciding what goes into the final work. Each repository says this in its README.
🇵🇱 Student studiów II stopnia na Politechnice Gdańskiej (chemia obliczeniowa i bioinformatyka). Powyżej są projekty studenckie (chemoinformatyka, QSAR/QSPR, lipidomika, modelowanie procesów, statystyka) oraz projekty poboczne. Kod powstawał z pomocą narzędzi AI. Każde repozytorium ma opis po polsku.