University of Dundee

In the past decade, biological sciences have witnessed a major shift towards data-driven research. Consequently, high-performance computing has become a standard research tool in life sciences. Biological data, however, is not only large but it is highly complex. This complexity requires alternative approaches to conventional data analysis. Machine learning (ML) has emerged as a powerful methodology that can successfully tackle the analysis of complex biological data. It is a hopeless task to attempt to develop a mathematical model of an elephant. Yet, a three-year-old child can with ease point at an elephant in a photo. The child was shown a picture of an elephant and told that the object in the picture was an elephant. In other words, she learnt to recognise an elephant by seeing photos of it and now she can identify it on her own. ML emulates the learning process on a computer. Instead of building a precise description of patterns, the computer is "taught" to recognise them. This is a paradigm shift from conventional computing and, thus, has its own challenges. Notably, the brain is well suited for learning by example, yet it performs poorly when it comes to long divisions. Computers, on the other hand, have been designed to perform numerical operations with great speed and precision. It is, therefore, not surprising that emulating an inherently heuristic process such as learning on a computer would require a substantial computational effort. With the recent advent in Graphics Processing Unit (GPU) and Solid State Drive (SSD) technologies, the necessary computer power has become broadly available. It is, however, not surprising that traditional High-Performance Computing facilities are not well-suited for ML applications. ML has been successfully used in biology for more than two decades. An excellent example is the prediction of the viability of cancer cells when exposed to a drug. The idea is to associate a response (e.g., whether a cancer cell survives or not) to a set of characteristics or features (e.g., which genes were mutated and what chemical properties of the drug are). In the so-called supervised learning, the machine is presented with a large set of training data that contains correct responses for given input parameters. Based on that data, the machine learns to predict the response for new, previously unseen parameters. A major challenge is that it is often not easy to identify what the appropriate features are. Cells are very complex and it is often unclear which are the most relevant features that determine a specific response, e.g., mutations of which genes one should consider, etc. An expert is, therefore, required to prepare the appropriate training set. In recent years, so-called deep learning techniques have revolutionised the learning process by allowing the machine to automatically extract the key features from raw data. This is achieved by a set of model neurons, inspired by biological neural cells, organised in a layered network (i.e., a neural network). The information propagates through layers of the network, which enables each layer to capture more and more abstract features in the data. This drastically reduces the need for carefully tailored training sets and makes the ML applicable to a wider range of problems, especially those where expert-made training sets are not available or too costly to make. Deep learning ML approaches, however, require substantial computational resources. Typical deep learning neural networks contain tens to hundreds of layers, thousands of neurons, and hundreds of thousands of links between them. Training them, therefore, requires hardware that operates at TFLOPS speeds (trillions of operations per second) and can access the data at several GB/s. The aim of this proposal is to build a designated GPU-based system for applying deep-learning ML methods in fundamental biological research at the University of Dundee.

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.

Doctoral Training Partnerships: a range of postgraduate training is funded by the Research Councils. For information on current funding routes, see the common terminology at www.rcuk.ac.uk/StudentshipTerminology. Training grants may be to one organisation or to a consortia of research organisations. This portal will show the lead organisation only.

It is well-known that identical living cells within complex multi-celluar organisms (e.g. humans) can respond to environmental signals and perform different, but co-ordinated roles. One of the most striking examples of this is in embryo development. This process is due to ?cell differentiation? and until recently, it was thought this behaviour was only found in complex, multi-cellular organisms. However, recently it has been discovered that simple single-celled organisms such as bacteria, also display cell differentiation and so to some extent can behave as multi-cellular collectives . One of the most striking examples of cell differentiation in bacteria occurs in the formation of biofilms. A biofilm contains billions of individual cells encased in a self-produced polymer glue. Despite each bacterial cell being genetically identical, the community soon differentiates into sub-populations, each carrying out a different role. Just how this complex multi-cellular decision making process occurs is far from understood. Almost all bacteria that occur in the natural environment live in these closely-knit biofilms and they are important to all aspects of our lives e.g. human health, the effective treatment of sewage and even daily dental care: plaque is a bacterial biofilm. The overall aim of this work is to better understand how differences in the way genes within individual bacterial cells respond to environmental signals leads to the segregation of these cells into sub-populations, each of which behaves in an entirely different, but apparently co-ordinated, manner. In this project I will focus on the well-studied bacterium Bacillus subtilis, which is used to produce enzymes for cleaning products (e.g. biological washing powder) and has growing potential as an alternative and environmentally friendly pesticide. Biofilms are so complex and cover such a wide range of scales (it would take 1000 cells laid end-to-end to cross a pin head but biofilms can be centimetres or even metres across) that it is necessary to take an inter-disciplinary ?systems? approach. I intend to use a combination of powerful modern mathematical modelling techniques supported by state-of-the-art molecular biology experimental procedures and will work closely with biofilm expert Dr Nicola-Stanley Wall, University of Dundee. Very recently, I appear to have uncovered an entirely new mathematical theory that may explain cell differentiation. It is this exciting discovery that motivates the work of this proposal.

Scotland has the highest number of drug related deaths (DRDs) per capita out of any European country (National Records of Scotland, 2021). The majority of DRDs in Scotland occur due to opioid induced respiratory depression (OIRD) (Information Services Division ISD - NHS National Services Scotland, 2018). OIRD is caused by opioid drugs decreasing the sensitivity of the brain stem to rises in CO2, which can lead to rapid respiratory failure. Naloxone is effective at overdose reversal; however, its administration requires bystander presence. 62.9% DRDs in Tayside occurred when individuals used substances alone (Tayside Drug Death Review Group, 2019). Preventing overdoses in people who use drugs (PWUD) alone is therefore an unmet need. This project is an ongoing mixed-methods observational cohort study aiming to investigate whether an accelerometer sensor attached to the chest can accurately and reliably capture respiratory patterns of PWUD to determine trigger points for an emergency response during an overdose. The study aims to assess the acceptability of the device to PWUD and first responder stakeholder groups from the third sector support groups and first responders to create an intervention pathway. The planned study duration is 12 months, from January 2022 to January 2023.