Skin Microbiome
The skin microbiome comprises bacteria, fungi, viruses, and other microorganisms. It acts as a protective barrier against harmful pathogens, helping prevent infections and strengthening the immune system. Dysbiosis of the skin microbiome is associated with various skin conditions such as eczema, acne, and psoriasis. The microbiome is a complex and dynamic ecosystem consisting of intra-species and inter-species interactions and bi-directional interactions between host and microbiome. Using our computational modelling expertise, we aim to understand how interactions within the skin’s ecosystem help maintain microbiome symbiosis, or cause dysbiosis, and discover whether and how we can alter the microbiome dynamics to advance personalised medicine and dermatological therapies.
Eczema Pathogenesis
Eczema is an itchy, chronic inflammatory skin condition that affects ~20% of children with a high socioeconomic burden. While the first line of eczema treatment is topical anti-inflammatory therapy, targeted biologic therapies have become increasingly popular for moderate to severe patients since the approval of dupilumab in 2016. We have been developing several computational models to understand eczema pathogenesis towards personalised medicine – the right treatment for the right patient at the right time. Our computational models describe the dynamic interplay between the skin barrier, environmental stress, and immune dysregulation across multi-spatial-temporal scales. Perturbations triggered at one part of the system can propagate to another part through interactions. Using our computational modelling expertise, we aim to entangle the system-level mechanisms underpinning the transition of disease phenotypes to help design treatment strategies for personalised medicine.
EczemaNet
Eczema is an itchy, chronic inflammatory skin condition that affects ~20% of UK children with a high socioeconomic burden. Timely and adequate treatment is crucial to manage eczema symptoms to avoid worsening. Eczema severity assessment is critical for deciding which treatments are needed and if specialist referral to secondary care is required. However, judging eczema severity on 2D images is challenging within the increasingly popular field of teledermatology. AI-powered skin image analytics can propose a promising solution. We have been developing EczemaNet, an AI-powered computer vision pipeline for automated assessment of eczema severity from digital camera images. EczemaNet first uses a convolutional neural network to detect regions of interest from an image, followed by probabilistic predictions of the severity of seven clinical disease signs. We aim to develop EczemaNet as a trustworthy and reliable software as a medical device.
EczemaPred
Eczema is an itchy, chronic inflammatory skin condition that affects ~20% of UK children with a high socioeconomic burden. Eczema symptoms manifest as relapses and remissions that are often unpredictable, making treatment difficult and increasing patients' burden. Predicting future disease states for individual patients will help design personalised treatment strategies because eczema symptoms fluctuate dynamically in a highly heterogeneous manner, and treatment responses often differ between and within patients. We have introduced EczemaPred, a computational framework to predict patient-specific dynamic evolution of eczema severity based on statistical machine learning. Using our data-driven modelling expertise, we aim to better understand eczema severity dynamics and comorbidity, help design personalised treatment strategies, and provide patients with streamlined descriptive, predictive, and prescriptive information.
Pre-school Wheeze & Asthma
About half of pre-school children (under 5 years) experience at least one episode of wheezing, with breathlessness and difficulty in breathing. One-third of the wheezers develop asthma, the commonest long-term condition affecting children in the UK. Given life-long impact of childhood asthma on lung health, we need to identify who requires interventions to prevent progression to asthma. Using our computational modelling expertise, we aim to elucidate the biological mechanisms that lead to the development and progression of asthma from pre-school wheezing. We are specifically interested in the impact of environmental exposures on lung function, early-life immune response to allergens and viruses, and the role of lung epithelial barrier and airway microbiome in progression to asthma.
Systems Mycology
Fungal infections affect over a billion people worldwide and remain a global threat to human health. The invasive fungal infection is associated with unacceptably high mortality rates even with treatment. Antifungal drugs are available, but with the issue of antifungal resistance. There is a real need for new antifungal drugs to be developed based on understanding of the dynamically changing host-pathogen interactions. Using our computational modelling expertise, we aim to develop tools for more informative data collection for the Mycology community, addressing a potential lack of standardization of data collection inter- and intra-labs. Concerted efforts to obtain dynamic data collection will enable development of mathematical models that can help achieve mechanistic understanding of the therapeutic effects towards development of new antifungal drugs.
