Nicolas Shiaels
DPhil Student in Biophysics, University of Oxford
About
I am pursuing a DPhil in Biophysics based in both the labs of Professor Achilles Kapanidis (group's website) in the Physics department of the University of Oxford and Dr Luke Clifton (website) at ISIS Neutron and Muon Source at the Rutherford Appleton Laboratory. My research focuses on the development of novel, detection methods for pathogens such as viruses and bacteria and understanding their surface membrane and proteins and their importance to infection using single particle microscopy techniques and neutron reflectometry.
Short Bio
I was born and raised in Cyprus in a small village, called Lemba, just outside Pafos. I completed my high school education at the Lyceum of Emba, accompanied with afternoon lessons to complete my A-level studies. In 2012 aften graduating school I joined the National Guard of Cyprus to complete my mandatory military service. Having been selected to become an officer I attended the Military Enginnering School of Loutrakion. On completion of my officer training I became a platoon leader and by the end of the second year of my service I was promoted to second-lieutenant in the reserves. I studied Physics at the University of Oxford at St Hugh's College and I am now pursuing a DPhil in Biophsyics at the same university and I am now a member of St Anne's College. In my free time I enjoy exercising, watching anime series and this year I am attempting to learn Japanese.
Funding and Scholarships
Viruses/COVID-19 Efforts
Every year millions of people are affected by viral disease, which in many cases can be fatal. For example, influenza causes mild epidemics every winter, but new emerging strains occasionally result in severe pandemics of worldwide proportions such as the current COVID-19 pandemic; these pose an imminent threat to humanity as they have wiped out millions of people in the past and the current pandemic has forced us to change our way of life and devastated the economy. Thus, a rapid, straightforward and sensitive method for detecting viruses will result in the quicker and more accurate treatment of patients, and save lives.
A novel method of detecting Viruses: 1min Summary
We have invented a new method of detecting intact viruses in 1-5min! In October 2020 the University of Oxford released a press release, which attracted the interest of both the academic world and the general public. If you would like to rad more about the work please refer to the relevant pre-print.
COVID-19 JR Hospital Oxford
In the past few months we have taken our test from the lab to the clinic, and are currently working with clinical collaborators at the John Radcliffe Hospital to validate the assay on COVID-19 patient samples. Excitingly, the combination of rapid labelling (Publication) and deep learning classification allows the detection of SARS-CoV-2 in less than 5 minutes, significantly faster than existing diagnostic tests.
Machine Learning
Applying a wide range of machine learning classification algorithms including k-nearest neighbour and deep convolutional neural networks (CNN) on biological systems.
Deep Learning CNN for Virus Classification
i
Deep learning has consistently proven to perform better than other machine learning techniques especially when it comes to image classification. The reason we chose to utilise this method is because of the large amount of data we can easily gather with our experimental assay described above and the fact that we wouldn’t need to know the specific features used for classification. The algorithm itself should in principle be able to identify the most suitable features and fine tune them to achieve optimal classification performance. Conventional analysis methods have proven ineffective when it comes to diffraction limited images of viruses and it is impossible to distinguish them by eye. Since we wanted to differentiate different viruses, we used supervised learning essentially forcing the algorithm to identify the unique features of each virus strain.
Neutrons
Neutron and X-ray reflectometry are powerful tools for the analysis of the structure at interfaces. The data from these techniques describes both the roughness and composition of the interface between two bulk phases as as information about the structure at or close to this. The interfacial structure is encoded in Kiessig fringes, an interference pattern which produces continuous oscillations in the experimental data. The frequency and magnitude of these oscillations are directly related to the thickness of a given layer of material at an interface and the difference in scattering magnitude of this layer from its surroundings.
ISIS Neutron Training Course 2020
The anual ISIS Practical Neutron Course is a fully-funded, competitive programme aimed at PhD and post-doctoral researchers who have little experience of neutron scattering, but whose future research aims to make use of neutron scattering techniques at ISIS. I followed the soft matter stream of the course, learning about sample preparation and deuteration, small-angle neutron scattering and neutron reflectometry. The course was very well organised and a lot of fun, and I made friends with some really intelligent and fun people! For more information about the course click here.
Teaching
2020-2021
Intellectual Property and Publications
Intellectual Property
- Method of Diagnosing a Biological Entity, and Diagnostic Device.
Reference number: N419195GB
Filing date: 27th April 2020
Publications
2020
- Shiaelis N. et al.Virus detection and identification in minutes using single-particle imaging and deep learning medRxiv 2020.URL
2017
- Cording, A.,Shiaelis N. et al. Targeted kinase inhibition relieves slowness and tremor in a Drosophila model of LRRK2 Parkinson’s disease. npj Parkinson's Disease 3, (2017).URL