Breast cancer is the most common cancer in women in the UK, but how angiogenesis is regulated in breast cancer is unknown. The growth of blood vessels into tumours (angiogenesis) is essential for cancer growth, and anti-angiogenic therapy is an important new strategy to treat cancer. Interestingly, women with Down Syndrome very rarely develop breast cancer, possibly as a result of the presence of three copies of chromosome 21. We will determine how three copies of chromosome 21 suppresses breast cancer angiogenesis. I will focus upon the involvement of a specific chromosome 21 molecule, termed JAM-B, which we already have evidence for a role in angiogenesis. I will use both clinical material and experimental mouse models. This research will help us to understand the disease better and possibly lead to the development of therapies to inhibit breast cancer progression.

In this project, our main goal is to design a generative deep learning system, which can compose contemporary classical music. Composition approach in this study is comprised of two main subparts, namely generating symbolic music and transferring symbolic music into the sonic/audio domain. Interpretable continuous control over musical attributes is one of the goals here as it is an open problem in the field and improves the functionality of such systems. We challenge the idea of learning for perfection, which is a typical approach in machine learning, and offer unusual-yet-compelling musical material aligned with contemporary classical music in the hope of expanding the current aesthetics of music and contributing to the musical culture with fresh and innovative compositional ideas. Primarily, this system is designed as a co-creative composition tool. Secondarily, there might be some other potential use cases such as robotic musicianship applications in the future.

Manually transcribing music is labour intensive, but still the dominant way of creating data to train an automatic transcription system (in the hope that one day it could replace manual transcription). One reason why the manual process is so time-consuming, is that annotations are typically made for entire music pieces, even for sections that in hindsight contribute little to the improvement of an automatic system because the previous version of the system already managed to successfully transcribe that section. The underlying cause is that the current state-of-the-art in music transcription, regardless of the exact characteristic that's being transcribed (e.g. tempo, melody, instrumentation, harmony), is based on deep learning techniques, which are not very good at representing uncertainty. The field of active learning aims to include uncertainty as part of the training process, such that an iterative workflow can be established where the segment that is deemed most informative gets presented for transcription first. In this project, the leading active learning approaches will be adapted to one or more music transcription tasks. The result will be integrated into a browser-based transcription tool, which will subsequently be used to study the difference in personal preferences and subjectivity between transcribers.

Genetically modified mice have been used to identify a protein in skin cells that protects them from developing cancer. By investigating how this protein protects these cells, we aim to understand what role it might have in human skin cancer. Fibroblast growth factors (Fgfs) play fundamental roles in biology, influencing how cells behave during development, maintenance and repair of the body. They act by binding to specific receptors, one of which, Fgfr2b, is the protein that has this protective role. Mice lacking Fgfr2b in their skin cells are very susceptible to skin cancer and the aim of this project is to identify why such cells are so prone to developing into tumours. As well as being relevant to cancer research, Fgfr2b is also implicated in developmental abnormalities, thus in addition to addressing questions of fundamental biological interest, our research will address clinically significant problems. Hopefully, as a result of these studies, new drugs might be designed to stimulate the beneficial actions of these receptors and to have an impact on cancer development in man. My group consists of one post-doctoral fellow and one PhD student, and we are all interested in various aspects of Fgf biology. This project grant would fund an additional post-doctoral fellow who would be able to spend three years using our mouse model of skin cancer, together with cell culture approaches, to dissect out the mechanisms by which Fgfr2b protects cells from becoming cancerous.