PhD position in AI-driven Discovery of New Cancer Genes

As a PhD student, you will:

• Retrain and adapt sequence-based deep learning models (e.g. Borzoi) to prioritize non-coding variants with downstream functional effects
• Integrate AI-based predictions with large-scale cancer whole-genome and (single-cell) RNA-seq datasets
• Identify non-coding regulatory regions enriched for functional somatic mutations across cancer types, with a focus on rare cancers
• Validate predictions using somatic eQTL analyses and gene expression data
• Contribute to open-source software and publish results in peer-reviewed journals
• Present your work at international conferences and consortium meetings
• You will receive close supervision and training in both machine learning and functional genomics, and work in a highly collaborative, interdisciplinary environment.

• Required: Holds (or will soon obtain) a Master’s degree in Bioinformatics, Artificial Intelligence, Computational Biology, Data Science, or a related field
• Has solid programming skills in Python, and experience developing or modifying analysis pipelines or machine-learning models
• Is motivated to work at the interface of AI and genomics
• Has strong analytical thinking skills and enjoys working with large datasets
• Has good written and spoken English

• Nice to have: Experience with deep learning models
• Familiarity with genomics concepts (eQTLs, gene regulation, RNA-seq, WGS)
• Experience working on HPC systems or cloud computing
• Interest in cancer biology
• Previous experience in an international research environment and strong references from direct supervisors are considered an advantage

This position is best suited for candidates who enjoy working independently on open-ended research questions and who are comfortable combining machine learning with biological interpretation. The Franke group values open scientific discussion, frequent interaction, and independence. PhD students are expected to actively present unfinished work, ask questions, and contribute to collaborative problem solving.