Cancer Research UK Manchester Institute - Principal Computational Biologist

Job Title:

Principal Computational Biologist

Category:

Other

Closing Date:

24/11/2024

• Machine Learning & Bioinformatics
• Salary £38,000 - £45,000 per annum (depending upon experience)
• Job Ref: MI/24/45_2
• Duration: Fixed term until 31^st August 2027

Background:
The National Biomarker Centre (NBC), in the Cancer Research UK Manchester Institute, is an acknowledged world leader in the study of minimally invasive biomarkers, or liquid biopsies, for improving cancer management. We have a highly interdisciplinary team working to develop a blood test to identify the tissue-of-origin (TOO) of tumours for patients diagnosed with a cancer of unknown primary (CUP). CUP is a metastatic cancer for which the primary tumour cannot be determined despite extensive diagnostic work up, limiting treatment options for these patients. We have shown in a proof-of-concept study that a machine learning classifier, named CUPiD, can determine the TOO of cancers using cell free DNA (cfDNA) methylation profiles generated using our in-house method T7-MBD-seq^{1, 2}.

About the role:
We are pleased to be able to offer the opportunity for a Principal Computational Biologist to join the NBC, to work on the optimisation and validation of our TOO classifier¹. You will work alongside our multidisciplinary team of clinicians, molecular biologists, and computational scientists on a Cancer Research UK funded project to develop CUPiD towards a clinically validated approach.

The main area of focus will be the development and application of statistical/machine learning approaches to enhance classifier performance, for example through:

• Refinement of data augmentation strategies for generating training data sets;
• Multi-omic data integration, combining genomic and fragmentomic features alongside methylation profiles;
• Optimisation of classifier hyperparameters;
• Improved statistical models for normalisation of our genome-wide methylation sequencing data and for detection of differentially methylated regions.

The optimised classifier will then be validated in a large cohort of cfDNA samples from patients with known cancer types. This is an exciting opportunity to develop and apply computational methods to high-throughput data in a translational research project with a clear line of sight to clinical application. 

¹Conway, Pearce, Clipson et al. Nature Communications (2024);           ²Chemi, Pearce et al. Nature Cancer (2022). 

About you:
You should have a PhD in Computational Biology/Bioinformatics, Statistics, Computer Science (or related discipline), or a relevant postgraduate degree in Computational Biology/Bioinformatics, Statistics, Computer Science or related discipline plus significant relevant experience. You will have significant experience in the development and/or application of statistical/machine learning methods, particularly supervised learning approaches, and demonstrable experience in areas of bioinformatics pertaining to the analysis of high-throughput data. You will also have significant experience in writing code for robust and reproducible analysis. Experience with generative machine learning models, Bayesian models or generalised linear models for count data is desirable, as is an understanding of liquid biopsies, cancer genomics/epigenomics, and cancer biology.  

You will have excellent communication skills and apply these skills to interdisciplinary collaborations. Experience of multidisciplinary teamwork would be beneficial.

Why choose the CRUK NBC?
The Cancer Research UK National Biomarker Centre is a leading and highly specialised translational research centre within The University of Manchester core funded by Cancer Research UK, the largest independent cancer research organisation in the world.

How to apply?
To apply for this position please visit our website.

For any informal enquiries about this post, please contact Dr Alex Clipson: alexandra.clipson@cruk.manchester.ac.uk