Code for performing data analyses of proteomics data from persons with multiple sclerosis (MS) and healthy controls. Protein levels were measured, using proximity-extension assay combined with next-generation sequencing (Olink Explore), in cerebrospinal fluid samples and plasma samples from persons with MS (n = 186) and healthy controls (n = 43). A machine learning approach were used to identify protein biomarkers which could predict diagnosis, disease activity and long-term disability outcomes in multiple sclerosis.