Cell2Sample: PhD position for multi-omics computational framework

Achieving precise patient care requires an accurate understanding of the mechanisms underlying disease development and drug resistance, but these processes are often heavily dependent on complex interplays between cells within their microenvironment. Cellular heterogeneity can be attributed to cell type composition in the microenvironment, activation states, and their function in terms of interactions with other cell types (e.g., immune cells and tumour cells; immune cells and microglia). While advanced single-cell molecular and imaging data can address some aspects of this complexity, their availability is limited due to high costs and cannot be expanded to large series of patient samples, which is necessary for biomarker discovery. Furthermore, these techniques individually cannot cover all modalities (e.g., proteome) as efficiently as others (e.g., transcriptome), which makes it challenging to provide a complete picture.

The Cell2Sample project aims to overcome these limitations by establishing a comprehensive computational framework that integrates diverse omics data types — including (single-cell) RNA/DNA sequencing and spatially resolved molecular data — available from both public sources and ongoing projects by ADORE researchers.

Would you like to know more about the different phases within the PhD trajectory? You can read more about this on this page.

We are looking for a PhD student who will focus on bulk multi-omics data integration, working in close collaboration with another PhD student appointed for spatial transcriptome analysis in the Bioinformatics Lab of the Department of Epidemiology and Data Science.

The core objective of this project is to enhance Statescope — a Bayesian framework developed by our team for cell state profiling using bulk RNA-seq data. This enhancement will focus on integrating additional omics layers, such as proteomics and phospho-proteomics, to characterize individual cell types based on their multi-omics profiles and pathway activation states. Expanding Statescope to incorporate diverse omics modalities aims to improve deconvolution accuracy, especially for datasets lacking single-cell data for certain modalities. This expansion will require the development of tailored statistical models for each data type.

We are looking for a PhD student with the following profile:

• MSc in Bayesian statistics, biomedical informatics, bioinformatics, computer science or a closely related discipline;
• Experience with computer programming (R/Python/C/C++ or etc) and work-flow management (e.g., Snakemake);
• Basic understanding of or strong interest in genomics / transcriptomics / proteomics data analysis;
• Excellent communication skills and fluency in English, both written and spoken.