This seminar series serves as a platform to share fundamental advancements, methodological breakthroughs, and innovative applications at the intersection of pharmaceutical research, applied mathematics, and computational sciences. By bringing together domain specialists, this seminar aims to uncover common threads, challenge existing methodologies, and foster collaboration and cross-pollination of ideas.
Past Seminars
Leveraging Machine Learning and Molecular Generation for Next-Generation Drug Design
Speaker: Martin Šícho, Department of Informatics and Chemistry, University of Chemistry and Technology Prague, Czech Republic
Time: 12 April 2024 13:00
Location: lecture hall C
Abstract: In the pursuit of efficient drug discovery, integrating computational methods with experimental approaches has become imperative. Therefore, this talk explores a number of approaches that leverage machine learning models to complement the work of medicinal chemists. Beginning with site of metabolism prediction, which involves modeling atom reactivity with atom descriptors and decision trees, we take a look at the possibilities of predicting regioselectivity of both phase I and phase II metabolic enzymes. Furthermore, we also visit the field of de novo drug design, where we employ computational techniques to design novel compounds with desirable properties. By combining docking simulations with molecular generation and reinforcement learning, we streamline the drug design process, prioritizing readily obtainable molecules with interesting binding properties. Additionally, we introduce the simple concept of localized combinatorial libraries, which enable efficient exploration of chemical space for de novo drug design. Many of these concepts will be presented alongside practical case studies with critical assessment of their strengths, limitations, and potential impact on accelerating drug discovery efforts in the future.
Digital twins for liver function evaluation and metabolic phenotyping – The role of cytochrome P450 variability
Speaker: Matthias König, Institute for Theoretical Biology, Humboldt-University Berlin, Germany
Time: 15 January 2024 13:30
Location: large meeting room of the dean's office
Abstract: Accurate assessment of liver function remains a major challenge in hepatology. Dynamic liver function tests, which non-invasively measure liver function and metabolic phenotyping in vivo, hold great promise in addressing this challenge. These tests assess liver function by monitoring the clearance of specific test compounds, thereby providing insight into the metabolic capabilities of the liver.
Our study uses comprehensive whole-body physiologically based pharmacokinetic (PBPK) models to simulate these dynamic assays, encompassing absorption, distribution, metabolism and excretion (ADME) processes. PBPK models are essential for investigating the metabolism of drugs and their systemic effects. This research highlights our use of PBPK models as digital twins for metabolic phenotyping and liver function assessment. We have established the first open pharmacokinetics database, PK-DB (https://alpha.pk-db.com), which contains curated data from over 600 clinical trials, to develop and validate our models [1-6].
Our models are individualizable and stratifiable, allowing simulation of lifestyle factors and co-administration effects on drug metabolism. We have applied our models to various clinical questions, such as simulating individual outcomes after hepatectomy using an indocyanine green model and investigating the influence of CYP2D6 gene variants using a dextromethorphan model integrated with drug-gene interactions. These models are hierarchically structured, describing metabolic and other biological processes in organs such as the liver and kidneys, and linked to whole-body physiology. All models and data are accessible for reuse in a reproducible manner, encoded in the Systems Biology Markup Language (SBML) [7-8].
A critical aspect of these models is the role of cytochrome P450 enzymes, which are central to the first phase of drug metabolism. We investigate how variations in CYP enzymes affect drug metabolism, taking into account factors such as pharmacogenomics, enzyme quantity variability, and spatial distribution within the liver (hepatic zonation). Our research includes: (1) creating a database of CYP protein levels, examining influences such as age, sex, body weight, smoking, and alcohol consumption (https://protein-distribution.streamlit.app/); (2) investigating the effect of CYP2D6 genotypes on dextromethorphan pharmacokinetics [5]; (3) analyzing species differences in zonation patterns of key CYPs [9]; (4) investigating how spatial zonation patterns of CYPs affect drug metabolism and potential harm (SPT model).
This presentation will provide an overview of PBPK models and discuss the impact of CYP variability on drug metabolism.
- PK-DB: pharmacokinetics database for individualized and stratified computational modeling. Grzegorzewski J, Brandhorst J, Green K, Eleftheriadou D, Duport Y, Barthorscht F, Köller A, Ke DYJ, De Angelis S, König M. Nucleic Acids Res. 2021 Jan 8;49(D1):D1358-D1364. doi:10.1093/nar/gkaa990. pmid:33151297
- Pharmacokinetics of caffeine: A systematic analysis of reported data for application in metabolic phenotyping and liver function testing. Jan Grzegorzewski, Florian Bartsch, Adrian Köller, and Matthias König. Frontiers in Pharmacology 2022, Vol12. doi:10.3389/fphar.2021.752826. pmid:35280254
- Prediction of survival after hepatectomy using a physiologically based pharmacokinetic model of indocyanine green liver function tests. Adrian Köller, Jan Grzegorzewski, Michael Tautenhahn, Matthias König. Front. Physiol., 22 November 2021. doi:10.3389/fphys.2021.730418. pmid:34880771
- Physiologically based modeling of the effect of physiological and anthropometric variability on indocyanine green based liver function tests. Adrian Köller, Jan Grzegorzewski and Matthias König. Front Physiol. 2021 Nov 22;12:757293. doi:10.3389/fphys.2021.757293. pmid:34880776
- Physiologically based pharmacokinetic (PBPK) modeling of the role of CYP2D6 polymorphism for metabolic phenotyping with dextromethorphan. Grzegorzewski, J., Brandhorst, J., König, M. Front Pharmacol. 2022 Oct 24;13:1029073. doi:10.3389/fphar.2022.1029073. pmid:36353484
- A physiologically based pharmacokinetic model for CYP2E1 phenotyping via chlorzoxazone. J. Küttner, J. Grzegorzewski, HM. Tautenhahn, M. König. bioRxiv 2023.04.12.536571 (preprint). doi:10.1101/2023.04.12.536571
- SBML Level 3: an extensible format for the exchange and reuse of biological models. SM Keating, D Waltemath, M König, F Zhang, A Dräger, C Chaouiya, FT Bergmann, A Finney, CS Gillespie, T Helikar, S Hoops, RS Malik-Sheriff, SL Moodie, II Moraru, CJ Myers, A Naldi, BG Olivier, S Sahle, JC Schaff, LP Smith, MJ Swat, DT, L Watanabe, DJ Wilkinson, ML Blinov, K Begley, JR Faeder, HF Gómez, TM Hamm, Y Inagaki, W Liebermeister, AL Lister, D Lucio, E Mjolsness, CJ Proctor, K Raman, N Rodriguez, CA Shaffer, BE Shapiro, J Stelling, N Swainston, N Tanimura, J Wagner, M Meier-Schellersheim, HM Sauro, B Palsson, H Bolouri, H Kitano, Akira Funahashi, H Hermjakob, JC Doyle M Hucka, and SBML Community members. Mol Syst Biol. 2020;16(8):e9110. doi:10.15252/msb.20199110
- The Systems Biology Markup Language (SBML): Language Specification for Level 3 Version 2 Core. M. Hucka, F. Bergmann, C. Chaouiya, A. Dräger, S. Hoops, S. Keating, M. König, N Le Novere, C. Myers, B. Olivier, S. Sahle, J. Schaff, R. Sheriff, L. Smith, D. Waltemath, D. Wilkinson, F. Zhang. J Integr Bioinform. 2019 Jun 20;16(2);. doi:10.1515/jib-2019-0021.
- Cross-Species Variability in Lobular Geometry and Cytochrome P450 Hepatic Zonation: Insights into CYP1A2, CYP2E1, CYP2D6 and CYP3A4 Mohamed Albadry, Jonas Kuettner, Jan Grzegorzewski, Olaf Dirsch, Eva Kindler, Robert Klopfleisch, Vaclav Liska, Vladimira Moulisova, Sandra Nickel, Richard Palek, Jachym Rosendorf, Sylvia Saalfeld, Utz Settmacher, Hans-Michael Tautenhahn, Matthias König*, Uta Dahmen* (* equal contribution) bioRxiv 2023.12.28.573567 (preprint). doi:https://doi.org/10.1101/2023.12.28.57356
Insight into lipid-based systems by molecular dynamics simulations
Speaker: Markéta Paloncýová, Czech Advanced Technology and Research Institute, Palacký University Olomouc, Czech Republic
Time: 21 November 2023 13:00
Location: large meeting room of the dean's office
Abstract: Lipids play a fundamental role in cellular processes, serving as structural components, regulators of substance flux within cells and the human body, and influencing enzyme function. This talk explores the field of molecular dynamics (MD) simulations in lipid studies, offering simultaneous atomistic and sub-picosecond resolution of dynamic processes. The journey begins with simple models of cell membranes, investigating the impact of lipid composition on structure, fluidity, and permeability. Progressing to drug-membrane interactions, we dive into biased simulation methods tailored for this purpose. The exploration extends to simulations of cytochromes P450 (CYPs), membrane-embedded enzymes, revealing how MD simulations unveil the influence of membrane composition on CYP structure and dynamics, as well as drug permeation through flexible enzyme channels. Finally, we scrutinize the capabilities of MD simulations to assess the role of lipids as delivery systems in RNA-carrying lipid nanoparticles (LNPs) at a coarse-grained resolution within microsecond timescales. This lecture aims to provide an overview of MD simulation applications in lipid systems, including a discussion of the limitations inherent to individual methods.
Is Parkinson's Disease the Scurvy of the 21st Century? (hydrogenish) Models of tinkering what's broken and breaking what's tinkered
Speaker: Bartek Lisowski, Department of Pharmaceutical Technology and Biopharmaceutics, Faculty of Pharmacy, Jagiellonian University Medical College, Poland
Time: 26 September 2023 13:30
Location: lecture hall C
Abstract: The burden of neurodegenerative diseases is increasing worldwide. In the United States alone, more money is spent each year on studying their mechanisms and possible therapeutic interventions than was spent on the entire Manhattan Project. However, despite these investments, no significant breakthrough has been made in the last 30 years. The best humanity can do is to find novel ways of supplementing what has been lost in the process of neuronal death, such as dopamine in the case of Parkinson’s disease. This is one of the similarities between neurodegeneration and scurvy, but it is not the only one. During the talk, we will examine at least two more common aspects of these terrible conditions - one seemingly from the past and the other casting a shadow on the future. This analogy will provide a context for the critical discussion of the utility of computational and theoretical models in understanding neurodegeneration and helping to develop better ways to assist people who suffer from it. We will look back in history to inspire the hopes for the future. In his groundbreaking 1913 work, Niels Bohr made no mention of uranium, much less plutonium. However, he devoted a lot of attention to the hydrogen atom - the simplest of all. And although the Danish physicist did not know it yet, his idealization buried (unfortunately, to some extent also literally) the old world and created a new one - the same one in which we still live today. Can something as complicated as the brain - or even more complicated: a dying brain - be understood through considerations written in pencil on a piece of paper, when the most powerful microscopes and tomographs are not up to the challenge?
Well, if we've already done it with an atom...
Physiologically Based Pharmacokinetic (PBPK) models – what, why, how, when, and where, all in 60 minutes!
Speaker: Sebastian Polak, Department of Pharmaceutical Technology and Biopharmaceutics, Faculty of Pharmacy, Jagiellonian University Medical College, Poland
Time: 25 September 2023 13:30
Location: lecture hall C
Abstract: Over the past two decades there has been tremendous rise in the use of physiologically-based pharmacokinetic (PBPK) modeling in the area of drug discovery, development and to support SUPAC changes. PBPK models combine and integrate information on the biological system of interest (i.e., morphophysiological parameters and their distribution in various populations) and drug properties (i.e., physico-chemical properties, binding, clearance, and formulation information) within a user defined trial design (i.e., dose amount, type and duration of application, number of subjects). Such an approach allows to simulate alternative hypotheses on the nature of interplay between the above listed factors, based on the ‘what if’ questions. The increased use of PBPK models throughout the drug development process is reflected in successful new drug application submissions to the regulatory bodies with the request to waive clinical trials and replace them with the PBPK based simulation results. This includes the use of PBPK models at the regulatory level for various reasons: supporting alternative bioequivalence (BE) approaches, e.g., virtual BE studies, defining a safe space for critical attributes, extrapolating bioavailability predictions and BE assessments from healthy to diseased populations and many others.
The aim of the lecture is to present the scientific background and discuss current achievements of the PBPK modelling. This includes recent case studies and discussion on limitations, obstacles, challenges, and future.
Pharmacological Models and Numerical Algorithms for the Refinement of Therapeutic Approaches for Retinal Diseases
Speaker: Elfriede Friedmann, Institute of Mathematics, University of Kassel, Germany
Time: 20 June 2023 16:00
Location: lecture hall A
Abstract: We present a model for drug distribution in the vitreous body that enables personalized therapy in ophthalmology. The standard treatment for age-related macular degeneration is anti-vascular endothelial growth factor (VEGF) drugs administered by repeated injections. The underlying model consists of a recently developed pharmacological model for the drug, which we extend by including spatial resolution, coupled with models describing the physiology of the vitreous medium. The resulting model is a system of partial differential equations, convection-diffusion-reaction equations coupled with Darcy-, Navier-Stokes- or viscoelastic Burgers flow. We are performing long-term three-dimensional Finite Element simulations for drug distribution in the human eye to gain new insights in the underlying processes using computational experiments. The developed simulations are used for the therapy optimization, for which specific output functionals are evaluated. We can perform accurate and efficient treatment testing, calculate the optimal injection position, perform drug comparison, and quantify the effectiveness of the therapy using the developed functionals.