Supervisors: Rolf I. Skotheim and Bjarne Johannessen
Prostate cancer is the most common cancer diagnosis among Norwegian men, presenting highly variable disease courses ranging from completely harmless to aggressively lethal. A persistent challenge in its clinical management is distinguishing between patients with indolent, slow-growing tumours who can safely undergo active surveillance and those with aggressive, potentially lethal disease requiring immediate radical therapy. This clinical decision-making is severely hindered by spatial tumour heterogeneity and multifocality, where approximately 75% of patients harbour multiple, separate malignant tumour foci that are clonally distinct and have completely separate mutational and gene expression profiles. Consequently, standard single-biopsy molecular tests suffer from severe spatial sampling bias.
To move past low-resolution DNA tests, we look to the "extra dimension" of the RNA transcript variants. Virtually all human genes undergo alternative splicing, producing a vast variety of transcript isoforms. In cancer, the splicing machinery is frequently dysregulated, resulting in highly cancer-specific RNA splice junctions (exon-exon junctions) and transcript variants that are absent in normal tissues.
For an RNA junction to be clinically useful as a diagnostic or prognostic biomarker, it must satisfy a critical prerequisite: high absolute expression, which is essential for maintaining analytical sensitivity, which is important particularly in clinical specimens like needle biopsies or blood samples where the tumour-derived RNA fraction is heavily diluted.
This project will utilize high-throughput whole-transcriptome RNA sequencing (WTS) data to systematically discover and validate highly expressed, cancer-specific RNA junctions. The student will develop a robust computational pipeline to facilitate the identification and prioritisation of highly expressed RNA splice junctions that meet the criteria of three strategic tracks:
Track 1: Absolute cancer-specificity (the diagnostic track)
Track 2: Prognostic and "lethal focus" enrichment
Track 3: Heterogeneity-agnostic prognostic enrichment
The MSc project is relevant for a student with strong programming skills in R and/or Python, with comfort with the Linux/Unix command-line, and who is able to work systematically.
What We Offer: Data and Infrastructure
Bulk tissue RNA-seq data is already generated and ready for analysis.
Secure data storage and high-throughput computational processing are fully supported by University of Oslo's secure Services for Sensitive Data (TSD) platform with allocated resources on Sigma2/NRIS supercomputer Linux clusters.
Interdisciplinary Research Environment: Our research groups are located at the Institute for Cancer Research at Radiumhospitalet, offering a highly collaborative environment where data science directly interfaces with clinical oncology.
Coursework in machine learning, statistics, or bioinformatics (e.g., expression analysis, statistical modelling).
An enthusiastic interest in utilizing computational and data-driven methods to solve real-world clinical challenges in cancer biology.
Interested candidates are encouraged to contact Rolf I. Skotheim (rolfis@uio.no) or Bjarne Johannessen (bjarnej@uio.no) with a brief description of their academic background, CV, and transcript of records.