Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
MSc thesis in deep representation learning for spatio-molecular data
Brain organoids recapitulate important features of human neurodevelopment and allow us to study this key developmental process at unprecedented scale and resolution. In the past, single-cell genomics technologies, including single-cell RNA-sequencing (scRNA-seq), have been used to study the molecular heterogeneity of brain development in organoid models. However, these assays dissociate the tissue and thus have limited capacity to learn about emerging spatial properties like regionalization or patterning. To overcome this limitation, the Treutlein lab applies highly-multiplexed spatial transcriptomic and proteomics technologies (e.g. MERSCOPE, 4i) to jointly measure spatial organization and molecular properties of organoids. Such datasets provide a unique opportunity to study the interplay between cellular communication, gene regulation, and spatial organization. However, a key challenge remains the integration of imaging and sequencing data across replicates, conditions, and time points to learn faithful cellular representations that can be mined for regulatory interactions. Thus, the aim of this thesis will be to develop new computational methods to infer meaningful cellular representations from rich and diverse spatial datasets.
Keywords: Organoids, brain development, deep learning, variational inference, optimal transport, deep representation learning
We're looking for a strong candidate to work on representation learning from spatial-omics data of organoid model systems. Organoids are state-of-the-art in-vitro systems to recapitulate multi-cellular behavior in realistic 3D model systems. Using advanced imaging and sequencing technologies, our lab acquires large volumes of high-dimensional phenotypic data across different conditions. For this thesis, we're interested in learning efficient representations from spatial and molecular properties. This is a computational project.
The lab: your thesis will be set at the Quantitative Developmental Biology lab at ETH Zurich's Department for Biosystems Science and Engineering (D-BSSE) in Basel, Switzerland.
Your qualifications: Proficient in Python, Experience with DL frameworks (e.g. pyTorch, JAX, PyG, pyro, ...), Interest in biological questions: development, regeneration, gene regulation, Background in mathematical/probabilistic modeling, Familiarity with Git, GitHub, CI, HPC
We're looking for a strong candidate to work on representation learning from spatial-omics data of organoid model systems. Organoids are state-of-the-art in-vitro systems to recapitulate multi-cellular behavior in realistic 3D model systems. Using advanced imaging and sequencing technologies, our lab acquires large volumes of high-dimensional phenotypic data across different conditions. For this thesis, we're interested in learning efficient representations from spatial and molecular properties. This is a computational project.
The lab: your thesis will be set at the Quantitative Developmental Biology lab at ETH Zurich's Department for Biosystems Science and Engineering (D-BSSE) in Basel, Switzerland.
Your qualifications: Proficient in Python, Experience with DL frameworks (e.g. pyTorch, JAX, PyG, pyro, ...), Interest in biological questions: development, regeneration, gene regulation, Background in mathematical/probabilistic modeling, Familiarity with Git, GitHub, CI, HPC
The aim of this thesis will be to develop new computational methods to infer meaningful cellular representations from rich and diverse spatial datasets. This includes benchmarking with existing methods and applying those methods to uncover novel insights into human brain development.
The aim of this thesis will be to develop new computational methods to infer meaningful cellular representations from rich and diverse spatial datasets. This includes benchmarking with existing methods and applying those methods to uncover novel insights into human brain development.
If this sounds interesting to you, please contact me at marius.lange@bsse.ethz.ch and include your CV & current transcript of records.
If this sounds interesting to you, please contact me at marius.lange@bsse.ethz.ch and include your CV & current transcript of records.