High Performance Computing

Computer-aided approaches are presently the world-leading techniques that allow for rationally managing the multilateral demanding process of pharmaceutical research. In this project, an arsenal of computational tools, together with our high-performance computing cluster Sofia will be utilized to discover and optimize new drug candidates for the treatment of various diseases (e.g., narcolepsy and Alzheimer’s disease). At first, we will employ classical approaches such as molecular docking and free energy perturbation using massive parallelization via a high-speed network to screen virtual ligand libraries. The main focus will be directed to the chemical modulation of biological targets that play an important role in neurodegeneration (e.g., monoamine oxidase A and B, amyloid beta binding alcohol dehydrogenase, cyclophilin D, cathepsins, superoxide dismutase 1, glycogen-synthase-kinases). In addition, advanced computational methods of quantum chemistry, MM/QM methods, metadynamics, bioinformatics, homology modeling, machine learning, data-mining, analysis of quantitative relationships between chemical structure and biological activity (QSAR), chemometrics, pharmacophore analysis, ADME/T prediction, in silico combinatorial synthesis with retrosynthetic algorithms will be utilized. All of these approaches will enable us to obtain a highly complex insight into efficient chemical modulation of the selected biological targets and to propose promising new drug candidates.
All neurodegenerative diseases are associated with various changes in the brain tissue which are not easy to discover. By nuclear magnetic resonance imaging, it is possible to obtain high-resolution grey-scale pictures of the brain that can be examined by machine learning methods to assist in correct early diagnosis. For these purposes, we will also utilize parallelized machine learning tools such as convolutional neural networks and support vector machines to create an intelligent diagnostic tool for the diagnostics of neurodegenerative diseases (e.g., Alzheimer’s disease, multiple sclerosis). The main focus will be centered on processing the data resulting from T1, T2, DWI, and FLAIR MRI experiments.