Proteins with non-trivial topology — knots, slipknots, and lassos — pose one of structural biology’s deepest questions: how does a linear chain reliably fold into a knot without getting permanently trapped? We build multidimensional energy-landscape models using coarse-grained Go-models and maximum-entropy approaches. Our central finding: topological barriers protect protein function and generate unique mechanical signatures measurable by single-molecule force spectroscopy.
We build the computational infrastructure for proteome-scale topological analysis — our databases KnotProt, AlphaKnot, LassoProt, and LinkProt catalogue topological complexity across hundreds of thousands of structures. Analysing AlphaFold predictions, we were among the first to discover new knot types in the human proteome, including the rare 6₃ stevedore knot never seen in experimental data. Direct Coupling Analysis (DCA) further extracts co-evolutionary signals to predict 3D residue contacts from sequence alone.
Classifying protein topology requires rigorous mathematics: we develop knot polynomials — Alexander, Jones, and HOMFLY methods — that uniquely identify knot types in open-chain biomolecules. We extend these tools to Seifert surfaces and genus theory, capturing topological complexity in both proteins and nucleic acids. Our open-source Topoly Python package makes these tools freely accessible to the community.
The TrmD enzyme — a tRNA methyltransferase essential for bacterial translation — owes its catalytic activity to a trefoil knot at its active site, a geometry impossible in unknotted proteins, making it a compelling antibiotic target with no structural human homologue. We apply DCA-guided computational docking and all-atom simulations to discover inhibitors for knotted proteins implicated in infectious disease, cancer, and neurodegeneration.
Physicists, mathematicians, biologists, and chemists — united by curiosity and rigorous thinking.
COST - EUTOPIA is the European network connected to the laboratory’s programme on protein entanglement. On this site it is grouped with the lab’s wider effort to classify entangled proteins and understand their free-energy landscapes.
The programme goes beyond classical knotted proteins by bringing complex lasso proteins and links into the same framework, connecting theoretical models with experimental validation.
This project focuses on entanglement in proteins: knots, slipknots, tadpoles, lassos, and links. These geometries are not visible in ordinary primary, secondary, or tertiary structure descriptions, but they can be essential for understanding folding, dynamics, and function.
The central goal is to understand the energy landscape of tangled protein structures through mechanical manipulation models. Theoretical predictions are designed to be supported and checked experimentally.
The main goal is to understand the role and function of knots in proteins. The project tests the hypothesis that knots and other non-trivial geometries, including slipknots, correlate with functional properties of proteins.
The work asks whether knot geometry is connected with active-site location, molecular motion, and co-evolving amino-acid pairs. Because knotted proteins are difficult to study experimentally, the project develops theoretical approaches as the first route to understanding their function.
This project studies proteins whose backbones form knots. Pulling a knotted protein by its termini would tighten the knot, while an unknotted protein would unfold into a chain.
The project asks how a single, topologically complex protein can perform biological work that might otherwise require several interacting parts. It focuses especially on understanding how knots affect function, using computer simulations and experimental techniques.
One motivation comes from methylation enzymes: in some organisms, reactions can depend on complex protein-RNA assemblies, while in others they are carried out by a single enzyme whose knotted topology may be functionally important.
The aim is to use theory from mathematics and physics, especially the geometry of curves and surfaces in three-dimensional space, to create tools for studying lasso-like protein topologies.
The project focuses on uncovering the folding process of recently discovered protein structures whose topology resembles a lasso.
This project characterizes the relationship between protein structure and function in proteins with lasso topology.
The work develops ways to describe lasso motifs and their functional implications, supporting the broader effort to detect, classify, and analyse lasso proteins.
The goal is to find selective inhibitors for the knotted TrmD tRNA methyltransferase.
The first step is to synthesize and test three ligands previously suggested as selective by AstraZeneca. The project is carried out in collaboration with the Jacek Jemielity group.
This project investigates the influence of non-trivial topology on ligand binding in the TrmD family.
TrmD proteins are characterized by a knot that forms an active site, which means topology can directly affect biological function through interactions with ligands. The work is carried out in cooperation with experimentalists in the Ya-Ming Hou group at Thomas Jefferson University, who determined structures of proteins with ligands and tRNA.
This project develops methodology for detecting macrocyclic links in proteins.
The work supports automated topological classification and analysis of linked protein motifs, helping make complex protein link topologies easier to identify and compare.
The aims are to develop new methods for automatically detecting link topology in protein structures, perform topological classification, and create a publicly available server and database: LinkProt.
The project also studies folding, determines free-energy landscapes of linked proteins, and searches for the influence of topology on protein function and properties. Simulations are supported by mathematical tools from knot theory and the theory of minimal surfaces.
