The use of high throughput data measured through the use of microarrays has started a new era in clinical diagnostics and prognosis. Microarrays, which are currently used to measure DNA content and DNA methylation, enable the measurement of tens of thousands of variables in a single experiment. Generally, the situation may then be described as having a large set of covariates that can be used to predict survival outcome. Typically such genomic data are high dimensional and consequently variable selection procedures such as Lasso and elastic nets are commonly used. However, these traditional variable selection procedures do not account for known, hierarchical relationships between data types. To take into account the differences between data types, it is proposed to explore different variable selection methods. There is a need to understand the relationships between genomic data and clinical data and observations in clinical studies. Such data will typically comprise patient outcome, including, for example, response, tumour size and growth kinetics, progression-free survival and overall survival. In addition, data will also be available for adverse events, including duration and severity. Relating genomic data to the various forms of patient data on efficacy and safety is of importance in order to understand more about patient selection for individual therapies and potentially prognostics for the outcome. AstraZeneca (AZ) has collected different types of genomic data in a number of clinical studies for several drugs in development for the treatment of cancer. The genomic data collected include copy number variation, somatic mutation, methylation, mRNA, miRNA and protein expression data. The clinical studies, which include Phase, 1, Phase 2 and Phase 3 studies comprise patients with various tumour types including ovarian cancer, non-small cell lung cancer and other solid malignancies. AZ have also identified suitable clinical studies or real-world evidence (RWE) databases in haematological and/or non-small cell lung cancer indications, which involve the collection of genomic data, and extract appropriate genomic data, together with longitudinal efficacy and safety data, all of which can be used in this project. In this context, RWE is generally understood to mean the information derived from electronic health records (EHR), which are used to centralise data across sites of healthcare and allow for population-level modelling and pharma-covigilance. It has been argued that research and randomised clinical trials at the site of care can generate results with superior external validity compared to conventional clinical trials, and facilitate the participation of a number of patients that, otherwise, would not benefit from the treatment. Difficulties in building prognostic models from genomics and real-world data (RWD) include interpretability of large datasets and inference of treatment effects. Aims The project, in collaboration with AZ, comprises the study of non-small cell lung cancer (NSCLC) data to develop prognostic tools, predictive models and reproduction of biomolecular pathways. The aims are summarised as follows: - Define predictive tools for time-to-event combining clinical data, which will include RWD and multicentre data, together with high-throughput genomic and transcriptomic data. - Analyse and test the biological plausibility for each gene pathway and biomarker described in the prognostic model. - Examine the predictive performance in RWD datasets and evaluate the feasibility of the prognostic tests with particular interest in improving false discovery rates from current techniques. - Analyse and undertake improvements in the protocols of clinical trials or observational cohorts from a disease discovery, data analysis, results-oriented viewpoint. The clinical usefulness of each prognostic model will be assessed in terms of its ability to predict clinical outcome in RWD

Elliptic curves are the heart of contemporary research in number theory. For example, their modularity played a crucial role in Wiles' proof of Fermat's Last Theorem. Understanding their rational points is the subject of the Birch and Swinnerton-Dyer Conjecture (BSD), one of the seven $1M Clay Mathematics Institute Millennium Prizes. This proposal is concerned with rational points on elliptic curves and linked to both BSD and modularity. The Darmon Programme is an ambitious project initiated by Henri Darmon, aiming to provide constructions of points on elliptic curves over number fields, and to prove new cases of BSD. The proposed action is an initiative to extend the Darmon Programme, make known constructions explicit and algorithmic to provide extensive data that will be invaluable to researchers in the field, and guide further theoretical work. Warwick has one of the largest and most active explicit number theory groups in the world, consisting of 2 professors, 3 lecturers, 8 research fellows and 12 PhD students, making it a natural host for the project. The supervisor Siksek, is a leading expert on elliptic curves, rational points and modularity, with considerable experience in supervising research including 5 Marie Curie fellows. Masdeu did his undergraduate studies in Barcelona. He completed his PhD (McGill) under the supervision of Darmon, and therefore has intimate understanding of the Darmon Programme. Masdeu then worked as Ritt Assistant Professor at Columbia University before joining Warwick as a postdoc in 2014. Therefore his research experience has almost entirely been confined to North America, even though his ambition is to establish himself as an independent researcher at a prestigious European university. Masdeu has substantial research results with 8 papers accepted in strong journals. This project will integrate him into the European research environment, and give him new skills and research directions to enable him to realize his goal.

Enacyloxin IIa is a polyketide antibiotic with activity against Gram-positive and Gram-negative bacteria that targets ribosomal elongation factor Tu. It has recently been identified as metabolite of Burkholderia ambifaria AMMD and shown to have clinically-relevant activity against Acinetobacter baumannii, a problematic pan-resistant Gram-negative pathogen. Despite its promising biological activity, enacyloxin IIa has not been used in the clinic, presumably due to stability issues. Preliminary experiments have provided evidence for an unusual mechanism of modular polyketide synthase chain release in enacyloxin biosynthesis, involving intermolecular condensation of an acyl carrier protein (ACP)-bound thioester with the C-3 hydroxyl group of (1R, 3R, 4S)-3,4-dihydroxycyclohexane carboxylic acid (DHCCA). The resulting intermediate undergoes epimerisation at C-1 of the DHCCA moiety. This project aims to explore the ability of the chain release enzyme to catalyse the acetylation of a variety of DHCCA analogues with an acetyl-ACP analogue of the polyketide thioester intermediate. It also aims to identify the enzyme responsible for epimerising C-1 of the DHCCA moiety. DHCCA biosynthetic genes will be deleted in B. ambifaria and enacyloxin analogues, with a stable amide group in place of the labile ester group and other modifications to the DHCCA-derived moiety, will be produced via a mutasynthesis strategy.

About 11M cases of cancer are diagnosed each year and there is a pressing need for new anticancer drugs. The world’s leading anticancer drugs are platinum-based compounds such as cisplatin. These types of platinum compounds are not targeted to cancer cells and also kill healthy cells. This can result in severe side-effects and the development of resistance, so that some patients stop responding to treatment. Methods for increasing the selective activation of platinum drugs at the site of the tumour are envisaged which address these problems. We have developed new platinum-based drugs which are non-toxic in the dark, but become highly toxic when activated by light. By shining light only on the area of cancerous tissue, we can target the activation of the drug to just the selected area. It is vital that we carry out more studies into how these drugs kill cancer cells, which will involve laboratory work and experiments to investigate if they can shrink/treat cancers. We hope that the wavelengths of light for activating these compounds make them suitable for treating surface cancers and that slight changes to the drugs will allow treatment of a wider range of cancers.

Abstracts are not currently available in GtR for all funded research. This is normally because the abstract was not required at the time of proposal submission, but may be because it included sensitive information such as personal details.