PhD student deep learning in genomics to decode cellular identities in the human brain

Description

The Yves Moreau and Stein Aerts research labs are looking for a shared PhD student to apply deep representation learning and large language models to decipher the genomic regulatory code with a focus on the human brain. You will engineer new AI models to predict celltype specific expression of all proteincoding genes in the genome across diverse cell types in different regions of the human brain. Techniques will include Transformers selfsupervised learning reinforcement learning and generative AI models. Explainability of your models is essential to discover new biological rules of the genomic regulatory code. As a secondary aim your trained models can be applied to interpret human genetic variation and to improve the prediction of disease risk from the genome sequence.

The project funded by an FWO Strategic Basic Research (SBO) grant will focus on designing and testing genomic enhancers to target transgenes to specific cell types in the mouse brain and in cultured human brain biopsies using adenoassociated viral vectors. Whereas you will focus on AI modeling other members in the consortium will perform wetlab experiments to test your predictions and designed DNA sequences. This provides an opportunity to get acquainted with wetlab experimental research. The PhD student will be stationed at ESATSTADIUS in the Moreau group but will be embedded in both the Moreau and the Aerts labs as well as the new VIB Center for AI & Computational Biology and the Leuven Bioinformatics and AI community. The PhD student will be jointly supervised by Yves Moreau and Stein Aerts.

Publications

Check out some of our recent publications

• From the Aerts lab related to deep learning in genomics:
  • Minnoye & Taskiran Genome Research 2020;
  • Kalender Atak & Taskiran Genome Research 2021
  • Janssens Aibar & Taskiran Nature 2022
  • Taskiran et al. BioRxiv 2023
  • For all publications see
• From the Moreau lab related to deep learning:
  • Verplaetse et al. Genome Biology 24 (1)
  • Passemiers et al. arXiv:2304.0
  • Oldenhof et al. arXiv:2303.0
  • Raimondi et al. Bioinformatics 37 (16) 22
  • For all publications see

Profile

• You obtained a Masters in Computer Science Artificial Intelligence Bioinformatics Physics Engineering Bioengineering or equivalent. Please note that to be admitted to the doctoral training distinction ( cum laude) based on your study results or professional realizations is a strict requirement.
• Comprehensive AI background (deep learning probabilistic modeling generative AI)
• Proficient in Python programming
• Experience with machine learning is a plus (e.g. PyTorch/Tensorflow/Keras)
• Experience with explainable AI (e.g. SHAP) is a plus
• Experience with highperformance computing software containers
• Experience with genomics is a plus but not essential
• Ability to work independently and in a team.
• Proficiency in oral and written English.

We offer

• Access to stateoftheart compute & GPU infrastructure
• A stimulating international research environment
• You can be engaged in our SBO consortium bringing together research labs across the University of Leuven University of Antwerp and VIB
• Competitive salary and benefits
• Fully funded PhD scholarship but encouraged to apply for a national PhD fellowships (e.g. FWO).
• Starting date: as soon as possible

How to apply

Please complete the online application procedure and include a detailed CV two reference letters and a motivation letter.

For further information and questions please send an email to Yves Moreau () or Stein Aerts ()