The Institute of Genetics and Molecular Medicine (IGMM; www.igmm.ac.uk/) at the University of Edinburgh has an exciting opportunity for an experienced and highly motivated Bioinformatician to play a core role in our genomics strategy. They will work from the new, state of the art Systems Medicine building at IGMM, designed around our thriving community of computational biologists. This post offers an excellent opportunity to work with a variety of distinguished IGMM scientists on a cutting edge cancer genomics project. The post would suit a postdoctoral biologist with bioinformatics experience or a postdoctoral computer scientist with a strong interest in biomedicine. The post in generously funded for 3 years and the deadline for applications is 20th October 2017.
Project:
There are over 600 new diagnoses of ovarian cancer in Scotland annually. High grade serous ovarian cancer (HGSOC) constitutes 75% of these cases and almost always presents with disease that has spread beyond the pelvis. Prognosis is generally poor and five year overall survival rate remains around 40%. HGSOC is characterized by ubiquitous somatic TP53 mutations and defects in DNA repair pathways, leading to particularly elevated rates of structural variants (SVs), thought to play critical roles in driving tumourigenesis. However, this disease is still poorly understood at the genomic level, and there are only modest numbers of samples available. To date the largest whole genome sequencing (WGS) study included only 92 tumour samples (Patch et al, Nature, 2015, 521:489).
Over the next two years we will process all data emerging from a large study (co-funded by AstraZeneca) of HGSOC in 200 Scottish patients at high (matched tumour 70X and normal sample 30X) coverage, providing unique opportunities to examine the origins and consequences of the somatic mutational landscape of HGSOC. We will relate this WGS analysis to previous WGS HGSOC data (Patch et al, Nature, 2015, 521:489), as well as a 450 patient Scottish formalin-fixed paraffin-embedded (FFPE) panel sequencing dataset with full prospective data and copy number analysis from another (Gourley lab) project. The Scottish samples are all unusually well annotated within the Edinburgh Ovarian Cancer Database (Gourley lab), with detailed histopathological and clinical data available. Building upon the primary analysis of WGS and panel sequencing data we propose three substantial projects to be pursued during this fellowship.
(i) Recent work on a cohort of 560 breast tumour patients has shown that the genome-wide mutational spectrum present in each tumour, ascertained from WGS data, can itself offer valuable diagnostic information, allowing more accurate therapeutic stratification (Davies et al, Nat Med, 2017, 23:525). For the first time we will establish methods to calculate similar mutational spectra in WGS HGSOC samples and explore their clinical utility in predicting clinical variables, including disease progression and therapeutic responses.
(ii) We will integrate our WGS data with previously published (Patch et al, Nature, 2015, 521:489) HGSOC WGS data and associated clinical data from our collaborators (Prof Sean Grimmond, University of Melbourne) to establish the generality of our observations. We will also test the clinical utility of mutational spectra calculated from the larger panel based FFPE sequencing cohort. Given the large collections of FFPE tumour samples worldwide the development of such analyses based upon (relatively inexpensive) panel sequencing may have broad translational impact across oncology internationally.
(iii) We will establish novel protocols at IGMM to identify complex SVs in HGSOC sequencing data, such as those arising from chromothripsis and chromoplexy, and define their utility for patient stratification. Through our involvement in the International Cancer Genome Consortium (Campbell et al, Nature, in review) we have access to a variety of recently developed (but often unpublished) tools and benchmarking data for the detection of complex SVs in tumour WGS data. There is intense interest in such SVs and their value as biomarkers of disease progression and relapse, but this has not been studied in HGSOC. With the largest HGSOC WGS dataset collated so far (see (ii)) we will be uniquely placed to detect candidate biomarkers, and test the additional predictive value of including complex SVs in tumour mutational spectra.
Essential:
• PhD in Biology, Bioinformatics, Statistics, Mathematics, Computer Science or an equivalent postgraduate qualification in a closely related discipline.
• Published research experience in statistical data analysis especially with applications in computational biology.
• Experience in handling large, complex datasets in compute cluster environments.
• Demonstrable skills in project management, in particular the management of multiple projects concurrently.
• Programming/scripting experience in Linux/UNIX environments.
• Strong analytical and statistical skills.
• Effective verbal and written communication skills.
• Good interpersonal skills and efficient time and project management skills.
Desirable:
• Experience in statistical programming languages especially R/Bioconductor.
• A background in cancer research and bioinformatics.
• Expertise in developing sequence analysis pipelines, statistical protocols, tools and databases for research in computational biology.
• Experience in variant databasing, variant visualisation/browsers and automated variant report generation.
• Familiarity with and published research experience in an important area of biomedicine.
• A good publication record in the analysis of high throughput sequencing data, or in other high-throughput biomedical data.
• Effective problem-solving and decision-making skills with the ability to troubleshoot complex analysis issues.
• Experience in communicating with scientists of diverse scientific backgrounds.
Please apply via the University of Edinburgh HR site:
www.vacancies.ed.ac.uk/pls/corehrrecruit/erq_jobspec_version_4.jobspec?p_id=041319