Protein design of novel immunotherapeutic drugs
antibodies - gene therapy - vaccines - cellular therapies - peptides
Schoeder Lab
at Leipzig University, Medical Faculty, Institute for Drug Discovery
Our Research Interests
We use computational protein design including machine learning to design proteins with new functionalities. Through interchanging application of computational protein design and experimental validation, we design a new generation of protein drugs.
The central paradigm of structural biology says that protein function is determined through protein structure, which is encoded in the protein sequence. While most of research has been done on predicting protein structure from protein sequence, protein design is considered the reverse protein folding problem.Computational tools we develop and study help us to design sequences that correspond to protein structures with the intended function.
We conduct computational protein design iteratively with experimental validation. As computational methods we use:
Protein are expressed either via bacterial or mammalian expression systems. Subsequently, they are purified and characterized using the following techniques and instruments:
- Rosetta
- Machine Learning (Language models, ProteinMPNN, ...)
Protein are expressed either via bacterial or mammalian expression systems. Subsequently, they are purified and characterized using the following techniques and instruments:
- Size exclusion chromatography, IMAC, IEX
- ELISA, SPR, BLI
- X-ray crystallography, biomolecular NMR
- nanoDSF, ITC
- activity assays
What we design:
check out: Schoeder and Schmitz et al.
- antibodies, Fabs, scFvs
- TCRs
- antibody-antigen interactions
- glycosylation sites
- vaccines, epitope-focused vaccines or prefusion stabilized glycoproteins
- viral capsid proteins
- peptides (for GPCRs)
check out: Schoeder and Schmitz et al.
Some of our projects:
Investigating the antibody response to AAVthrough antibody discovery and antibody design
Prediction of glycosylation sites and other post-translational modifications while designing proteins.We use artificial neural networks (ANNs) and integrate them in the Rosetta software suite.
Computational design of peptides targeting GPCRs.We are studying new methods for protein design in the context of GPCR drug discovery.
Funding of the lab
We thank RosettaCommons, the state of Saxony, the Faculty of Medicine of Leipzig University, the Coalition for Epidemic Preparedness Innovations (CEPI), the ScaDS.AI and the DAAD for their contributions to our research.
We are part of the SFB1423 "Structural Dynamics of GPCR Signaling". Check it out here: research.uni-leipzig.de/sfb1423/
Contact
Contact
clara.schoeder[at]medizin.uni-leipzig.de
Leipzig University Medical FacultyInstitute for Drug DiscoveryLiebigstr. 19D-04103 Leipzig