Projects Offered

1 PhD project offered in the IPP winter call Molecular Mechanisms in Genome Stability & Gene Regulation

Scientific Background

Protein-protein interactions (PPIs) are important mediators of most cellular processes. Understanding how proteins interact with each other, ideally at structural resolution, is key in understanding molecular processes, how disease mutations would disrupt PPIs, and in guiding design of binders with therapeutic potential. The ways by which proteins can interact with each other are highly diverse, yet this structural diversity is poorly understood. This is because experimental techniques to solve structures of protein complexes favor more stable interactions with larger interfaces and because we lack efficient algorithms to compute similarity between interaction interfaces across millions of comparisons. This hinders identification of novel modes of protein binding, i.e. those predicted by AlphaFold, and it hinders design of more balanced datasets of structurally resolved modes of binding to train the next generation of artificial intelligence (AI)-based structure prediction tools.

PhD Project: AI-assisted protein-protein interaction interface clustering and discovery of novel modes of protein binding

The Luck lab has substantial expertise in the study of modes of protein binding by experimental and computational techniques. We built pipelines based on AlphaFold and in-house developed software to predict structures of interacting proteins and in collaboration with the Steinegger lab, developed highly efficient AI-based algorithms to compute similarity between interaction interfaces. The PhD candidate will work on substantial improvements of these existing computational pipelines and use them subsequently to discover novel modes of protein-protein binding by performing structure prediction screens and by searching available resources of predicted protein complex structures. Experimental validation and characterization of potentially novel modes of protein binding is possible in collaboration with other members of the lab.

Desired (but not absolutely required) skills: programming in python, machine learning, and experience in protein structure analysis. 

Required skills: High motivation, curiosity, self-driven, critical thinking, strong team-player, good English, high interest in protein structure and machine learning.

This PhD project is funded via a Marie Curie Actions Doctoral Network Program called ProtAIomics. The PhD candidate should not have lived or worked in Germany for more than 12 months within the last three years to be eligible for the position in the Luck lab. The program will offer topic-related training within the network of research labs that are part of ProtAIomics and the student will participate in project-related stays in other labs. The position is well funded including coverage of all travel costs.

If you are interested in this project, please select Luck as your group preference in the IPP application platform.

Publications relevant to this project

Hubrich D, Alvarado Valverde J, Lee CY, Djokic M, Welzel M, Hintz K, Luck K (2025) Variant characterization in the intrinsically disordered proteome. BioRxiv, Link

Strom JM, Luck K. (2025) Bias in, bias out - AlphaFold-Multimer and the structural complexity of protein interfaces.Curr Opin Struct Biol, 91:103002 Link

Geist JL, Lee CY, Strom JM, de Jesús Naveja J, Luck K. (2024) Generation of a high confidence set of domain-domain interface types to guide protein complex structure predictions by AlphaFold. Bioinformatics, 40(8):btae482 Link

Lee CY, Hubrich D, Varga JK, Schäfer C, Welzel M, Schumbera E, Djokic M, Strom JM, Schönfeld J, Geist JL, Polat F, Gibson TJ, Keller Valsecchi CI, Kumar M, Schueler-Furman O, Luck K (2024) Systematic discovery of protein interaction interfaces using AlphaFold and experimental validation. Mol Syst Biol, 20:75-97 Link