Liquid biopsies have become increasingly important for disease diagnosis, but the focus has mainly been on nucleotides rather than proteins. Myotonic dystrophy type 1 (DM1) presents an opportunity to explore the use of protein analysis in liquid biopsies, as the disease causes significant tissue damage and protein leakage into blood. However, to achieve this, three challenges must be addressed: detecting low-abundance proteins, identifying their tissue of origin, and understanding functional perturbations. This will be achieved through four work packages: (i) develop performant machine learning models and optimising their features for data from the highly performant diaPASEF mode of the timsTOF instrument; (ii) use these features to create a sensitive and specific search engine for data analysis; (iii) tracing proteins back to their tissue of origin using tissue prediction models and perform functional protein association mining to create functional networks of proteins to elucidate functional perturbations; and (iv) combining the analytical advances with the biological context to obtain candidate biomarkers for DM1. This combination of enhanced analytics and relevant biological context will lead to more robust biomarkers and serve to demonstrate the utility of this approach for complex disease analysis and diagnosis.