— Computational deconvolution for patient stratification in the context of non-small cell lung cancer

Puzzle Freak

Computational methods can help to unravel the complexity of the tumor microenvironment of individual patients. We ask for your help to assess such methods to enhance our understanding of cancer heterogeneity and to improve patient classification.

Apply until Nov 30, 2020 / Xplorers Camp on Dec 11, 2020 

Question to be solved

Can patients be stratified by their immunological tumor environment and does this additional information improve classification models?

General Background

It has been shown that the levels of tumor infiltrating immune cells are associated with tumor growth, disease progression and patient outcome. Common methods for studying cell heterogeneity such as flow cytometry are limited by cell type marker selection and sample processing steps. Recent computational methods enable the prediction of cell type frequencies in tumor samples solely from gene expression data. An ideal method for feature engineering in the context of patient stratification.

Data Types & Technologies

  • Transcriptomic data from publicly available data sources such as the gene expression omnibus (GEO)
  • R packages: Caret, Cellmix

Supporting Material or Links

  • Shen-Orr, S. & Gaujoux, R. Computational deconvolution: extracting cell type-specific information from heterogeneous samples. Curr Opin Immunol 25, 571–578 (2013).
  • Newman, A. et al. Robust enumeration of cell subsets from tissue expression profiles. Nat Methods 12, nmeth.3337 (2015).

Needed Skills

  • Self-starter and organisational talent
  • Interest in cancer immunotherapy is a plus
  • Strong statistical background
  • Good knowledge in R or other statistical programming language


Jan Rieckmann
Team Lead Data Science

Form of Cooperation

Preferred scale: 3 months full-time (flexible models are also possible)

Possible format: Internship

How to present your Idea

Show us how you would approach the problem. You can prepare your idea proposal in 3-5 slides, any other idea/format to share your proposal with us is also welcome. We do not expect a bullet-proof solution to the problem.

By sending this to us via the submit button you agree to the following:


  • you confirm that you are the author of the submission and entitled to dispose of rights of use and exploitation of the contents of your submission, and that you have not yet granted any rights of use and exploitation to third parties that would be infringed by your submission;
  • you grant to Roche Diagnostics GmbH the unrestricted, sublicensable and exclusive right  to use and exploit your submission by all means known today or in the future. This includes without limitation the rights to reproduce, distribute, and exhibit your submission, as well as the right to communicate your submission to the public. You also grant to Roche Diagnostics GmbH the right to edit the submission, to translate it, and to create abbreviations and summaries (abstracts); the aforesaid rights to use and exploit also apply to such edited versions, translations, abbreviations and summaries. 
  • Roche Diagnostics GmbH will designate you as the author of the submission, and will recognize and respect your moral rights in the submission.


Any problems with the submit button? Please send your submission to healthcare.xplorers@roche.com.


Further information on our privacy policy can be found here.


Any problems with the submit button?
Please send your submission to healthcare.xplorers@roche.com.


Further information on our privacy policy can be found here.