Institute
The Institute of Cancer Research London www.icr.ac.uk, is one of the world’s most influential cancer research organisations, with an outstanding record of achievement dating back more than 100 years. The Institute of Cancer Research (ICR) discovers more new cancer drugs than any other academic institution globally, and with The Royal Marsden runs one of the world’s largest phase I clinical trial units for cancer. It is also a world leader in cancer biology and genetics, and in developing new forms of high-precision radiotherapy. Its mission is to make the discoveries that will defeat cancer.
Collaborators: Dr. Syed Haider, Prof. Andrew Tutt
Location: Remote
Project Requirements
- Practical experience applying multivariable regression modelling to large datasets
- Understanding of Cox regression for survival analysis
- Experience with stochastic k-TSP desirable but not essential
- PhD/MSc in STEM discipline
- Programming skills in R
- Bioinformatics knowledge useful but not essential as training will be provided
Project Summary
Cancer is a complex and heterogeneous disease with a far greater diversity of molecular underpinnings than phenotypic consequences. This molecular diversity is frequently exploited to identify biomarkers of patient outcome [1] and therapeutic response [2].
Over the past two decades, a large number of biomarkers have been identified, in particular by utilising RNA datasets. However, their application remains limited. A key factor underlying this limitation is ad hoc pre-processing and (co)normalisation steps required prior to their use, and hence preclude their application to single samples.
To circumvent this limitation, single sample predictors such as k-TSP have been successfully implemented [3]. Using gene expression profiles, k-TSP offers a rank-based decision system which generalises to new datasets however small the cohort is and remains independent of profiling platforms. With the emergence of pathway-based biomarkers [4,5], the challenges previously faced by gene-level biomarkers remain to be addressed.
In this study, we will test the appropriateness of k-TSP on pathway dysregulation scores. Pathway dysregulation scores will be transformed to rank-based decisions, resulting in a single sample predictor based on composite pathway dysregulation scores. Classification by the single sample predictor will be tested for association with clinical outcome. The performance of the single sample predictor will be compared to existing pathway-based biomarkers [5]. This study will utilise previously published breast cancer mRNA and outcome data [5].
References
[1] van ‘t Veer, LJ. et al. (2002) Gene expression profiling predicts clinical outcome of breast cancer. Nature
[2] Severson, TM. et al. (2017) The BRCA1ness signature is associated significantly with response to PARP inhibitor treatment versus control in the I-SPY 2 randomized neoadjuvant setting. Breast Cancer Research
[3] Tan, AC. et al. (2005) Simple decision rules for classifying human cancers from gene expression profiles. Bioinformatics
[4] Ben-Hamo, R. et al. (2020) Predicting and affecting response to cancer therapy based on pathway-level biomarkers. Nature Communications
[5] Haider, S. et al. (2018) Pathway-based subnetworks enable cross-disease biomarker discovery. Nature Communications