Sanofi iTech Award

Funding: | Sanofi-Aventis Recherche & Développement |
Runtime: | [01/2024 - 01/2025] |
Staff: | Malte Holmer |
Explaining protein ligand complexes with deep learning based scoring functions

Funding: | Sanofi-Aventis Deutschland GmbH |
Runtime: | [05/2023 - 04/2025] |
Staff: | Matthias Rarey, Tobias Harren |
CompLS - Runde 5 - Verbundprojekt: DrugSiderAI - KI-gestützte Methoden zur synergistischen Erforschung von Krankheitssymptomen und Medikamentennebenwirkungen

Funding: | BMBF |
Runtime: | [03/2023 - 02/2026] |
Staff: | Matthias Rarey, Paula Kammer |
Improved Protein-Ligand Structure Modeling using SAR Data

Funding: | F. Hoffmann-La Roche Ltd |
Runtime: | [01/2023 - 12/2025] |
Staff: | Matthias Rarey, Jonathan Pletzer-Zelgert |
Shaping Chemical Fragment Spaces on the Exemple of BICLAIM

Funding: | Boehringer Ingelheim International GmbH |
Runtime: | [09/2022 - 09/2025] |
Staff: | Justin Lübbers |
SAFE: Synthetically Accessible Fragment Space Extensions by Machine Learning-Based Approches

Funding: | DFG |
Runtime: | [05/2022 - 04/2025] |
Staff: | Malte Korn |
Verbundprojekt: Röntgenbasierte Medikamenten Entwicklungsplattform

Funding: | BMBF |
Runtime: | [04/2022 - 03/2026] |
Staff: | Thorben Schulze |
Center for Data and Computing in Natural Sciences

Topic: | In the natural sciences, research in many areas is based on the collection and analysis of large, complex and, in recent years, drastically increasing amounts of data. This is due to the increasing automation of experimental research, the digitalisation of complex technical systems and the growing importance of simulations. These challenges can only be met by closely linking the natural sciences with methodologically oriented research in computer science and applied mathematics. For this reason, the Center for Data and Computing in Natural Science (CDCS) is being established in the Science City Bahrenfeld (SCB) - covering science from basic and structural research to systems biology - in cooperation with the University of Hamburg (UHH), the German Electron Synchrotron (DESY) and the Hamburg University of Technology (TUHH). The CDCS initially consists of four application-focused, cross-disciplinary laboratories (CDLs), which are supported by an Informatics Core Unit (ICU). The CDLs focus on the following areas: Computational Astro and Particle Physics - Computational Photon Science - Computational Systems Biology - Computational Controls of Accelerators. The overall aim is to significantly strengthen the conditions for excellent research at the SCB in the field of computation. |
Funding: | State Research Funding 'Hamburg X' (funding code LFF-HHX-03 |
Runtime: | [12/2019 - 12/2024] |
Staff: | Matthias Rarey |
Partners: | Universität Hamburg, DESY, Hamburg University of Technology |
Method Development for Synthesis-aware Lead Optimization

Topic: | In the modern drug design process one of the main questions is how to optimize an initial lead compound towards a drug candidate with desired properties. The process of lead optimization is the modification of such an initial lead compound to enhance the desired pharmacological effect and minimize undesirable properties. During this optimization and modification phase, a major challenge is to ensure the synthesizability of the compound. The primary goal of this project is to develop computational strategies to address the problem of synthesis-aware lead optimization. Therefore state-of- the-art techniques, such as combinatorial chemistry and retrosynthetic analysis, are combined. |
Funding: | Bayer AG |
Runtime: | [10/2019 - 09/2022] |
Staff: | Uschi Dolfus |
Partners: | Bayer AG, Universität Hamburg |
DASHH - Data Science in Hamburg, Helmholtz Graduate School for the Structure of Matter

Topic: | DASHH is a newly established Helmholtz graduate school. In DASHH we harness data, computer and applied mathematical science to advance our understanding of nature. We aim to educate the future generation of data- and information- scientists that will tackle tomorrow’s scientific challenges that come along with large-scale experiments. |
Runtime: | [04/2019 - 09/2025] |
Staff: | Matthias Rarey, Michael Größler, Christiane Ehrt, Thorben Poburski |
Partners: | DESY, UHH, TUHH, HSU, Helmholtz Zentrum Geesthacht, HZI, MPI for the Structure and Dynamics of Matter, European XFEL GmbH |
Link: | https://www.dashh.org/index_eng.html |
GeoMine: Methoden zur geometrischen Mustersuche in Proteinstruktursammlungen

Topic: | Nowadays, more than 150 thousand biomacromolecular structures are available for life science research, a further steep increase can be expected in the next years. This wealth of data can be accessed and queried by all kinds of meta data, searching it by geometric arrangements and features is, however, a widely unsolved problem. Within GeoMine, we are developing database technologies for storing biomacromolecular structural data and quering it with geometrical patterns. |
Funding: | BMBF |
Runtime: | [04/2019 - 03/2022] |
Staff: | Konrad Diedrich, Joel Graef |
Interactive Pocket Exploration and Fragment Growing

Topic: | Goal of the project is to develop computational strategies tackling the problem of fragment growing in a binding pocket, using state of the art computer science methods. Emphasis is placed on the quick response of the developed system that will allow a user to interactively explore their chosen pocket. Eventually the system should also be capable of more advanced functionalities, such as water displacement or macrocyclization. |
Runtime: | [04/2018 - 03/2021] |
Staff: | Patrick Penner |
Partners: | Servier, Universitaet Hamburg (ZBH) |
Allianz protPSI - protein Pressure Specific Activity Impact: Modulation der Reaktivität von Proteinen und Thermodynamik durch Druck

Runtime: | [01/2018 - 12/2020] |
Staff: | Jochen Sieg |
Predictive Models: What's next?

Funding: | Merck KGaA |
Runtime: | [12/2017 - 11/2021] |
Staff: | Wolf-Guido Bolick |
Partners: | Merck KGaA, University of Hamburg (ZBH) |
de.NBI-Partner: Integrierte Webservices zur Unterstützung strukturbasierter, lebenswissenschaftlicher Forschung in der Protein-Biochemie, insbesondere Enzymologie

Topic: | Das Ziel des EnzymeStructures-Projektes ist es, Proteinstrukturdaten allen Wissenschaftlern im Bereich der modernen Lebenswissenschaften einfach zugänglich und nutzbar zu machen. Die Protein-Datenbank PDB hat sich bereits als weltweit zentraler Knotenpunkt für Experimentaldaten in diesem Bereich etabliert. EnzymeStructures wird die Strukturdaten mit wichtigen, automatisierten Präprozessierungsverfahren anreichern, die für die Nutzung der Daten in der biotechnologischen Forschung notwendig sind. Insbesondere werden Suchverfahren, die eine gezielte Auswahl forschungsrelevanter Proteine auf der Basis von Struktureigenschaften ermöglichen, entwickelt. Die Schwerpunkte liegen dabei auf der Betrachtung aktiver Zentren und Interaktionen zu niedermolekularen Substanzen wie Subtrate, Co-Faktoren und Inhibitoren. |
Funding: | BMBF |
Runtime: | [11/2016 - 10/2019] |
Staff: | Matthias Rarey, Stefan Bietz |
Partners: | de.NBI, AG Prof. Schomburg, Braunschweig |
Software: | http://proteins.plus |
Paper: |
|
Molecular Mind Map

Topic: | In der pharmazeutischen und biotechnologischen Forschung werden heute automatisiert große Mengen von Molekülen auf spezielle Eigenschaften experimentell getestet. Beispielsweise werden im High‐Throughput‐Screening Moleküle auf ihre Fähigkeit zur Inhibition spezieller Proteine untersucht. Die Auswertung dieser Daten erfolgt computergestützt. Häufig ist zunächst das Ziel, Gemeinsamkeiten und Unterschiede zwischen aktiven, d.h. positiv getesteten Molekülen zu identifizieren. In diesem Prozess ist Kreativität und Erfahrung von Wissenschaftlern ein ganz zentraler Baustein. Es gibt keine einfachen Schemata, in der Regel bedarf es der richtigen Intuition, um ein Modell zu erhalten, welches Aktivität durch die Struktur der Moleküle erklären kann. Für die computergestützte Analyse steht heute eine Vielzahl von Deskriptoren zur Beschreibung der Molekülstruktur zur Verfügung. Diese stellen die strukturelle und auch biologisch‐pharmakologische Ähnlichkeit der Moleküle bereits sehr gut dar. In der Regel sind es aber nur wenige ausgezeichnete Eigenschaften, deren Zusammenwirken Aktivität erklären kann. Aus diesem Grund werden bereits seit vielen Jahren maschinelle Lernverfahren zur Ableitung von Modellen eingesetzt. Dem Wissenschaftler kommt in diesem Prozess eine eher passive Rolle zu: Er kann mittels weiterer Testdaten die Qualität des Modells prüfen, aber nur in sehr geringem Maße konstruktiv in den Modellbildungsprozess eingreifen. Im Rahmen dieses Projektes möchten wir neue computergestützte Methoden entwickeln, um das intuitive Wissen des Anwenders mit maschinellen Lernverfahren in optimaler Weise zu koppeln. Der Grundgedanke ist, einen fließenden Übergang zwischen der händischen und maschinellen Klassifikation von Daten technisch zu ermöglichen. Dazu möchten wir neue, innovative, inkrementelle Lernverfahren wie beispielsweise Active Learning einsetzen. Bei diesen Verfahren werden die Daten schrittweise durch den Benutzer klassifiziert. Währenddessen können maschinelle Lernverfahren das Verhalten des Nutzers analysieren und Vorschläge zur Klassifikation unterbreiten. Der Nutzer kann anhand der Fehlerrate jederzeit entscheiden, zu einer dann automatischen Klassifikation überzugehen. Wichtig ist uns, den Ansatz so in Software umzusetzen, dass er sich für den täglichen Einsatz durch Naturwissenschaftler, nicht durch Informatiker, eignet. Die Software soll dazu graphischinteraktiv nach dem Vorbild von Mind Maps agieren. |
Funding: | Klaus Tschira Foundation |
Runtime: | [07/2015 - 06/2018] |
Staff: | Matthias Rarey, Rainer Faehrrolfes |
Partners: | University of Hamburg (ZBH) |
BioScore: New Methods for the energetical estimation of protein-protein and protein-ligand interactions for bioindustrial applications

Topic: | The project aims at further development of the Hyde scoring function to allow the estimation of interaction energies within proteins and of protein-protein complexes. These improvements enable a systematic optimization of enzymes and mutation analysis of protein complexes with regard to questions in the biocatalysis field. |
Funding: | BMBF (German Ministry of Science and Technology) |
Runtime: | [12/2012 - 11/2015] |
Staff: | Nadine Schneider, Agnes Meyder, Eva Nittinger |
Partners: | Bayer CropScience AG, BioSolveIT GmbH, University of Hamburg (ZBH), The project belongs to the Biocatalysis2021 cluster |
Link: | http://www.biocatalysis2021.org |
Software: | Hyde-Scoring |
Paper: |
|
An open, integrated and sustainable chemistry, biology and pharmacology knowledge resource for drug discovery (OpenPHACTS)

Topic: | Openphacts deals with the development of an open information platform for all kinds of data relevant for medicinal chemistry. The ZBH’s role is the joint development of innovative browsing technologies for biological and chemical data focusing on small molecules. |
Funding: | EU (European Union), IMI (Innovative Medicine Initiative) |
Runtime: | [03/2011 - 02/2014] |
Staff: | Katrin Schöning-Stierand, Matthias Hilbig, Tim Harder, Thomas Otto |
Partners: | Large academic-industrial consortium under the leadership of the University of Vienna |
Link: | http://www.openphacts.org |
Software: | ChemBioNavigator (CBN) (http://cbn.zbh.uni-hamburg.de) |
Paper: |
|
ViSoR-LeadIT: An integrated software platform for structure-based, virtual screening

Topic: | The execution of large-scale virtual screening campaigns is still a logistically demanding task. The ViSoR-LeadIT project aims at the development of an easy-to-use platform for process and data management related to virtual screening. The focus of our work is the development of methods dealing with protein flexibility preprocessing for structure-based virtual screening. |
Funding: | BMWI-ZIM (German Federal Ministry of Economics and Technology - Central Innovation Programme SME) |
Runtime: | [09/2010 - 08/2013] |
Staff: | Angela Henzler, Lennart Heinzerling, Stefan Bietz |
Partners: | BioSolveIT GmbH, c.a.r.u.s. HMS GmbH, University of Hamburg (ZBH) |
Fundamentals for Synthetic Biological Systems - SynBio

Topic: | This project deals with a detailed analysis of a biotechnologically highly relevant pathway for H2 production. The task of the ZBH group is the analysis of the structurally known key enzymes with respect to substrate specificity and feedback inhibition. |
Funding: | LEXI (State Excellence Initiative Hamburg), Joachim-Herz-Stiftung |
Runtime: | [01/2010 - 12/2012] |
Staff: | Karen Schomburg |
Partners: | Large academic consortium under the leadership of the Technical University of Hamburg |
Link: | http://www.tu-harburg.de/synbio/index.html |
Paper: |
|
Virtual Screening in the Space of Natural Products

Topic: | Structure-based virtual screening is widely used mostly to identify new lead structures among small synthetic compounds. This project aims at the optimization of virtual screening protocols for natural product libraries and its application to selected targets. |
Funding: | Beiersdorf AG |
Runtime: | [08/2008 - 07/2011] |
Staff: | Sascha Urbaczek |
Partners: | Beiersdorf AG, University of Hamburg (ZBH) |
Software: | Naomi |
Paper: |
|
ViSoR - Virtuelles Screening optimizing Reality

Topic: | ViSoR is a novel platform technology for managing experimental and virtual screening applications in academic research developed by c.a.r.u.s. HMS GmbH. The ZBH’s tasks are support for modeling the screening process and the integration of the TrixX virtual screening platform into ViSoR. |
Funding: | c.a.r.u.s. HMS GmbH |
Runtime: | [07/2008 - 03/2012] |
Staff: | Jochen Schlosser, Angela Henzler, Lennart Heinzerling |
Partners: | c.a.r.u.s. HMS GmbH, University of Hamburg (ZBH) |
Software: | TrixX |
Paper: |
|
Biocatalysis2021: COMPASITES: Computer-Aided Analysis of Protein Active Sites

Topic: | The project aims at new approaches to evaluate the druggability of active sites. Furthermore, the relationship between active site geometry and chemical properties on the one hand and protein function on the other should be elucidated in order to make functional predictions based on protein structure alone. |
Funding: | BMBF (German Ministry of Science and Technology) |
Runtime: | [06/2008 - 08/2012] |
Staff: | Andrea Volkamer |
Partners: | Merck KGaA, BioSolveIT GmbH, University of Hamburg (ZBH); The project belongs to the Biocatalysis2021 cluster |
Link: | http://www.biocatalysis2021.org |
Software: | DoGSite, DoGSiteScorer |
Paper: |
|
Combating Drug Resistance in CML and HIV-1 Infection

Topic: | The project is a joint effort for the development of new lead structures targeting the Deoxyhypusine pathway. The tasks of the ZBH are virtual screening and modeling calculations based on crystal structures of DHS (deoxyhypusine synthase). |
Funding: | BMBF (German Ministry of Science and Technology) |
Runtime: | [04/2008 - 12/2012] |
Staff: | Adrian Kolodzik, Björn Windshügel |
Partners: | University of Lübeck, Heinrich-Pette Institute for Virology, University Hospital Hamburg-Eppendorf (UKE), University of Hamburg (Chemistry Dep. and ZBH) |
Paper: |
|
GeoHYDE

Topic: | This project is the second in line with Bayer aiming at the development of a novel scoring function for protein-ligand complexes. The resulting function named HYDE is already quite well-known, see current literature. |
Funding: | Bayer Crop Science AG |
Runtime: | [01/2008 - 12/2010] |
Staff: | Nadine Schneider |
Partners: | Bayer Crop Science AG, Bayer Healtcare AG, BioSolveIT GmbH, University of Hamburg (ZBH) |
Software: | Hyde |
Paper: |
|
Computational methods to Support Structure-based Sequential Screening

Topic: | Sequential screening is the process of iteratively screening small compound sets and using this information for the selection of compounds to screen in the subsequent iteration. In this project, we jointly develop an approach for the visual analysis of affinity data within protein active sites as well as its use for target-driven scoring and docking methods. The concept of activity cliffs is central for this work. |
Funding: | N.V. Organon |
Runtime: | [07/2007 - 12/2009] |
Staff: | Birte Seebeck |
Partners: | N.V. Organon (now Merck & Co. Inc), University of Hamburg (ZBH) |
Software: | ISAAC |
Paper: |
|
Non-deterministic Search Algorithms for 3D De Novo Design

Topic: | The ZBH has an internationally unique reputation for the development of search methods for chemical fragment spaces. Within this project, a novel pharmacophore-based search engine is developed. Furthermore, novel tools for the analysis and prediction of bioactive conformations are created. |
Funding: | F. Hoffmann-La Roche GmbH |
Runtime: | [07/2007 - 12/2011] |
Staff: | Tobias Lippert, Christin Schärfer |
Partners: | F. Hoffmann-La Roche GmbH, University of Hamburg (ZBH) |
Software: | Qsearch, TorsionAnalyzer |
Paper: |
|
Two-dimensional representation of Protein-Ligand Complexes

Topic: | Protein-ligand complexes are usually visualized three-dimensionally. For the quick analysis of docking modes as well as for a simple exchange of structural data between scientists, the two-dimensional visualization method PoseView has been developed. PoseView creates pictures of complexes in text-book style fully-automatically. PoseView is meanwhile linked to the Protein DataBank (PDB). |
Funding: | Klaus Tschira Foundation |
Runtime: | [01/2007 - 07/2011] |
Staff: | Katrin Schöning-Stierand, Matthias Hilbig |
Partners: | University of Hamburg (ZBH) |
Link: | https://www.zbh.uni-hamburg.de/poseview |
Software: | PoseView |
Paper: |
|
Computational Methods for Focussed Library Design

Topic: | A frequent scenario in drug design is the creation of small focused libraries around initial screening hits. For this task, we develop a multi-criteria optimization method for a focused combinatorial library design named LoFT. LoFT considers physico-chemical compound properties as well as similarities and dissimilarities to known compound collections. |
Funding: | Boehringer-Ingelheim |
Runtime: | [11/2006 - 10/2009] |
Staff: | J. Robert Fischer |
Partners: | University of Hamburg (ZBH) |
Software: | LoFT |
Paper: |
|
c-Met-Screen

Topic: | Short-time virtual screening project aiming at the search for selective kinase inhibitors. |
Funding: | Pro-Qinase |
Runtime: | [02/2005 - 08/2006] |
Staff: | Birte Seebeck |
Partners: | ProQinase, University of Hamburg (Pharmacy Dep. and ZBH) |
Paper: |
|
Pre-Filters for Structure-Based Virtual Screening

Topic: | The project lays the foundation for our TrixX development, a novel structure-based virtual screening platform based on compressed bitmap index structures. |
Funding: | AstraZeneca |
Runtime: | [10/2004 - 03/2005] |
Staff: | Ingo Schellhammer |
Partners: | AstraZeneca , BioSolveIT GmbH, University of Hamburg (ZBH) |
Software: | TrixX |
Paper: |
|
NovoBench: A computational workbench for de novo drug design

Topic: | The aim of Novobench is the development of a toolset for de novo molecular design. The focus of the ZBH work is on the creation of structure-based tools for de novo design (FlexNovo) as well as the combination of de novo design with classical QSAR modeling (FragView and FragEnum). |
Funding: | BMBF (German Ministry of Science and Technology) |
Runtime: | [07/2004 - 06/2008] |
Staff: | Jörg Degen, Juri Pärn |
Partners: | BioSolveIT GmbH, University of Nürnberg-Erlangen, ALTANA Pharma AG, Lilly Forschung GmbH, 4SC AG, Molecular Networks GmbH, University of Hamburg (ZBH) |
Software: | FlexNovo, FragView, FragEnum |
Paper: |
|
Ligand-based 3D Scaffold Hopping

Topic: | A frequently occurring scenario is the necessity of redesigning a central part of a lead structure, the core of a molecule. The geometric fit of alterative cores plays a central role in this task. Within this project, we develop a database driven method for fast core replacement under pharmacophoric constraints named Recore, which is widely known and used today. |
Funding: | F. Hoffmann-La Roche GmbH |
Runtime: | [03/2004 - 02/2007] |
Staff: | Patrick Maaß |
Partners: | F. Hoffmann-La Roche GmbH , University of Hamburg (ZBH) |
Link: | http://www.biosolveit.de/recore |
Software: | Recore |
Paper: |
|
Development of an Adaptive Scoring Function for Structure-Based Virtual Screening

Topic: | Scoring can be considered the grand challenge problem of structure-based molecular design. In this project, a novel theory based on water network saturation was developed and converted into a scoring function named HYDE. HYDE is well known today, see current literature. |
Funding: | Bayer CropScience AG |
Runtime: | [09/2003 - 08/2005] |
Staff: | Ingo Reulecke |
Partners: | Bayer CropScience AG , University of Hamburg (ZBH) |
Software: | Hyde |
Paper: |
|