Latest Academic Works
Discovery of exact equations in mathematical databases
Equation discovery, closely related to and frequently discussed under the term symbolic regression, studies methods for identifying symbolic equations that describe relationships in observed data. Most existing approaches to equation discovery are designed for noisy measurements and therefore aim to discover approximate equations. In many mathematical settings, however, approximation is …
Localization in QSAR: theory of local model development, programming and implementation of the aiQSAR software package and the application of the methodology to the differential gene expression predictions
QSAR methods are a cornerstone of contemporary cheminformatics, yet conventional global models remain limited in their ability to provide reliable predictions for complex chemical spaces and endpoints. This dissertation develops, implements and applies a framework for localized QSAR modelling that attempts to address these limitations. The aiQSAR methodology is established …
Stress responses in grapevine: gene expression profiling under four stressors and genome editing approaches for phytoplasma resistance
Viticulture faces production challenges from multiple environmental factors, ranging from changing weather patterns and nutrient deficiencies to pathogen pressure. In this thesis, we deepened the understanding of how grapevine responds to water deficit, increased soil salinity, different aspects of iron deficiency, and infection by 'Candidatus Phytoplasma solani', by investigating transcriptomic …
Non-invasive end-of-line quality inspection of electric motors supported by artificial intelligence
Electric motors are produced in very high volumes and are critical parts of modern industrial and consumer systems, devices and appliances. Ensuring their reliability requires robust and automated End-of-Line (EoL) quality inspections that can detect mechanical and electromechanical defects. The EoL quality inspection is usually performed at the end of …
Biochemical and structural comparison of human and mouse cathepsin B complexes with darpins
Cathepsin B is a lysosomal cysteine protease implicated in various pathological conditions and therefore considered a potential therapeutic target. This study investigates the selectivity and binding of a designed ankyrin repeat protein (DARPin) 4m3 towards mouse cathepsin B (mCatB). It aims to explore the species-specific interaction of cathepsin B with …
The impact of stefin b and cystatin c deficiency on cancer progression and tumor microenvironment in a mouse breast cancer model
Lysosomal cysteine cathepsins participate in numerous physiological processes and are also involved in pathological conditions such as cancer and inflammation. Proteolytic activity of cysteine cathepsins is tightly regulated by endogenous inhibitors of the cystatin superfamily, among which stefin B is an intracellular, and cystatin C is an extracellular inhibitor. Altered …
Aerosol-deposited multicaloric thick films
Solid-state cooling technologies have received increased attention as suitable alternatives to conventional vapor-compression systems. In recent years, significant research has been done on caloric materials. In the pursuit of improved room-temperature caloric responses, with potential for multifunctionality, multicalorics have also been the focus of research, as these exhibit distinctive caloric …
Interplay of sensorimotor and reward learning in human motor control
Chemical space coverage remains a major challenge in analytical chemistry. Even with modern instruments and data-processing tools, only a small percentage of all existing organic compounds is detected, let alone identified. Broad range of molecular polarity is one of the main factors influencing this: the most hydrophilic and hydrophobic compounds …
Application of hydrophilic interaction liquid chromatography in environmental health and pharmacotherapy studies
Chemical space coverage remains a major challenge in analytical chemistry. Even with modern instruments and data-processing tools, only a small percentage of all existing organic compounds is detected, let alone identified. Broad range of molecular polarity is one of the main factors influencing this: the most hydrophilic and hydrophobic compounds …
Towards Robust Meshless Simulation of Fluid Flow: Adaptive Stabilisation and Refinement
Meshless methods are an increasingly popular alternative to traditional mesh-based numerical methods, offering the ability to handle complex geometries while avoiding the resource-intensive and difficult-to-automate process of mesh generation. Further benefits of meshless methods include dimension independence and natural support for spatially variable discretisation. The Radial Basis Function-generated Finite Differences …