
Dr. Mitya Pushkin, Dept. of Maths, York
Artificial Intelligence (AI), particularly deep learning models like AlphaFold, has revolutionised the prediction of protein structures from amino acid sequences. However, these models still struggle to predict the structures of collagen, the most abundant protein in multicellular organisms.Collagen forms a variety of fibrillar scaffolds crucial for tissue organisation and affects processes like bone mineralisation and cancer cell invasion and is promising for medical advancements. Recent progress in understanding collagen assembly has led to a reliable method for predicting its fundamental axially periodic structures, known as microfibrils. This project aims to leverage AI to analyze and design protein sequences capable of self-assembling into microfibrils. The student will take part in developing and training a deep learning model using this new methodology, based on thousands of mammalian collagen sequences. The goal is to generate and design novel collagen-like structural materials. The project offers an exciting opportunity for those interested in programming and passionate about developing mathematical models. It promises to be a great learning experience with the potential to make a significant impact on the field. Students will need to find their own accommodation and be expected to present their findings orally at a research day in York in September 2025.