Bio-medical knowledge bases are valuable resources for the research community. Original scientific publications are the main source used to annotate them. Medical annotation in Swiss-Prot is specifically targeted at finding and extracting data about human genetic diseases and polymorphisms. Curators have to scan through hundreds of publications to select the relevant ones. This workload can be greatly reduced by using bio-text mining techniques. Using a combination of natural language processing (NLP) techniques and statistical classifiers, we achieve recall points of up to 84% on the potentially interesting documents and a precision of more than 96% in detecting irrelevant documents. Careful analysis of the document pre-processing chain allows us to measure the impact of some steps on the overall result, as well as test different classifier configurations. The best combination was used to create a prototype of a search and classification tool that is currently tested by the database curators.