The analysis and statistical examination of biologically derived, often large-scale datasets, and the clear presentation of the results, necessitate the use of specialized software packages and computational tools for such tasks. In my work, I frequently use programs such as bwa, samtools, and GATK, which provide the bioinformatics framework for genomic studies, and I perform further analyses using custom scripts written in R or Python. These allow for the statistical validation of observed trends and the creation of striking, easily interpretable visualizations.

In the analysis of multidimensional datasets ("big data", e.g., mutation patterns from whole-genome sequencing, methylation data, transcriptomic profiles), I apply various, often computationally intensive methods and algorithms for data condensation and interpretation. These mostly involve different machine learning techniques from the areas of dimensionality reduction (PCA, UMAP, t-SNE), clustering (hierarchical, k-means), and prediction/classification (SVM, decision trees, regression).

One of the main focuses of my research is the analysis and biological interpretation of big data datasets generated by next-generation sequencing (NGS) instruments. The majority of the samples examined are derived from tumor tissues, and the primary goal is to answer how cancerous cells differ in their genomic properties from healthy ones, as well as to identify common characteristics shared by different types of tumors. The examination of molecular profiles enables the exploration of how modern, targeted therapeutic methods (e.g., PARP inhibitors, immunotherapy, etc.) could aid patient recovery and how effective they might be in treating a specific type of cancer.

In retrospective clinical studies with domestic and international partners, I focus on identifying biomarkers of malignant diseases and mapping their clinical significance. Primarily in lung cancer, I examine metastatic patterns, immunotherapeutic targets (PD-1/PD-L1), and the prognostic value of immune cell infiltration. My work includes developing novel statistical methods and visualisation techniques for evaluating complex clinical data.

In epigenetic ageing research, I develop mathematical models (epigenetic clocks) that estimate biological age based on DNA methylation. My goal is to create universal models that provide accurate results across different experimental platforms and tissue types. I work with both human and animal samples, developing algorithms applicable to alternative measurement methods (RRBS). The work extends to identifying differential methylation patterns.

I take part in projects that aim to develop public health monitoring methods through large-scale genomic data analysis. In my research, I have successfully drawn population genomic conclusions from datasets originally collected for mapping antibiotic-resistant genes in sewage samples, utilising trace amounts of human mitochondrial DNA. Through the analysis of millions of SARS-CoV-2 viral genomes, I detected rare evolutionary events (coinfection, recombination) and identified genetic changes affecting diagnostic method efficiency. My work contributes to demonstrating the public health applicability of genomic surveillance.