Postdoc Position Deep Learning for Glioblastoma Sequence-to-Function Models

Description

the VIB Center for AI & Computational Biology is a young research center dedicated to combining machine learning with in-depth knowledge of biological processes. Our mission is to study fundamental problems in biology and work towards foundation models of biological systems and innovative AI-driven biotech applications.

The Laboratory of Computational Biology in Leuven () led by Scientific Director Stein Aerts is seeking a talented postdoctoral researcher to develop next-generation sequence-to-function models for glioblastoma (GBM). Glioblastoma is the most aggressive form of brain cancer characterized by diverse and dynamic cell states that drive treatment resistance and poor prognosis.

In this project funded by the Foundation Against Cancer you will move beyond descriptive genomics to decipher the underlying regulatory logic of GBM. By leveraging single-cell multi-omics (scATAC-seq scRNA-seq) and spatial omics you will map enhancer landscapes at unprecedented resolution. The core innovation of your work will be integrating this data to train deep learning models that predict chromatin accessibility and gene expression patterns. These models will ultimately be used to design synthetic enhancers tailored to modulate specific GBM cell states offering a path toward highly targeted oncolytic virus therapies and immunomodulatory interventions.

Responsibilities

• Model Development: Build and train advanced deep learning architectures (e.g. CNNs Transformers Generative Models) to decode the regulatory logic of genomic enhancers in GBM and the tumor microenvironment.
• Synthetic Design: Use sequence-to-function models to design programmable synthetic enhancers capable of targeting specific cancer cell states or host cells.
• Data Integration: Integrate pan-cancer single-cell atlases with spatial transcriptomics to understand signaling pathways and gene-regulatory dynamics.
• Explainable AI (XAI): Ensure models provide mechanistic insights into cancer cell states moving from black box predictions to biological understanding.
• Collaboration: Work within a multi-disciplinary team and potentially engage with collaborators across Belgian universities.

Profile

• Education: PhD in Artificial Intelligence Bioinformatics Computer Science Physics Engineering or a related field.
• Programming: Proficient in Python.
• Machine Learning: Strong experience with frameworks like TensorFlow Keras or PyTorch.
• Preferred Skills:
  • Experience with Explainable AI (e.g. SHAP Integrated Gradients).
  • Familiarity with high-performance computing (HPC) and software containers.
  • Knowledge of cancer genomics or regulatory biology is a plus.
• Mindset: Ability to work independently while thriving in a collaborative international team.

We offer

• Cutting-Edge Resources: Access to state-of-the-art compute and GPU infrastructure including H100 and B300 GPU clusters.
• Innovation: The opportunity to apply a recently published proof-of-concept method for synthetic enhancer design to a critical real-world clinical challenge.
• Environment: A stimulating international research setting in a top-tier university.
• Funding: Minimum of 3 years of funding available; candidates are encouraged to apply for prestigious fellowships (EMBO MSCA etc.).
• Start Date: As soon as possible.

How to apply

Please complete the online application procedure via the VIB website and include:

1. A detailed CV
2. A motivation letter specifically detailing your interest in glioblastoma and deep learning
3. Two reference letters

For more information: