Applying Statistical and Deep-Learning Methods to the Joint Analysis of Pathology and Sequencing Data in Prostate Cancer

Project Description

There are many questions in Prostate Cancer that we can seek to address either through sequencing or imaging (e.g. tumour classification, prognosis, inference about the micro-environment). There are questions about sub-cellular localization that can be answered directly for small numbers of molecules via imaging approaches or inferred for large numbers of molecules from sequencing data. Further, it was recently shown that characteristics of the genome such as whole-genome duplication as well as focal mutations can be inferred from pathology images with the application of deep learning. While it seems likely that a portion of the RNA profile of a tumour could be similarly predicted from the pathology images. Considering data from the Cancer Research UK-funded Prostate Cancer ICGC Group, and the PanProstate Cancer Group, this project will look jointly to analyse pathology and sequencing data, both to understand better those data, but also to find the best estimates of characteristics that can be estimated from both.

Primary supervision will be provided by Prof Andy Lynch with secondary supervision from Dr Peter Caie, both within the School of Medicine. Prof Lynch is also a member of the School of Mathematics and Statistics, while Dr Caie is a platform lead at the Sir James Mackenzie Institute for Early Diagnosis, and the successful student, while based in the School of Medicine, will have interactions with these as well.

Andy Lynch was a PI on the phase of the Cancer Research UK-funded Prostate Cancer ICGC project that generated RNA- and Methylation sequencing data from 200 men with prostate cancer. These data joined the rich whole-genome sequencing and clinical data generated for the same men in phase I of the project. He is now involved in several analysis groups of the PanProstate Cancer Group project, an international effort to combine data from over 2000 prostate cancers, which is part-funding this project.

Dr Peter Caie is the PI of the Innovate UK-funded iCAIRD project in St Andrews. The iCAIRD project involves analysing thousands of pathology images through computer vision algorithms, developed in house, in order to perform automated and accurate diagnostics for gynaecological pathology. iCAIRD involves clinical and industry partners collaborating with our academic teams to ensure high impact of the research. Dr Caie’s other funded projects involve the analysis of the complicated cellular interactions and associated protein expressions within the tumour microenvironment, through automated image analysis, that may inform on disease aggression.

Funding Information

This is a 3 year funded PhD studentship comprising of tuition fees (Home/EU) and stipend at current research council rates.

Eligibility Requirements

Candidates should have a numerate background preferably with a master’s-level degree giving them expertise and experience in applied statistics or machine/deep learning.

Start date: September 2020 to January 2021, Negotiable.

Application Process

In the first instance, applicants should submit a CV and covering letter to Andy Lynch [email protected] and Peter Caie [email protected]

Closing date: Open until the position is filled

To apply for this PhD, please email

Before sending your email, please double check you have followed all guidelines in this listing and have included a reference number if asked to do so.