The fact that data is scattered over many tables causes many problems in the practice of data mining. To deal with this problem, one either constructs a single table by propositionalisation, or uses a Multi- Relational Data Mining algorithm. In either case, one has to deal with the non-determinacy of one-to-many relationships. In propositionalisation, aggregate functions have already proven to be powerful tools to handle this non-determinacy. In this paper we show how aggregate functions can be incorporated in the dynamic construction of patterns of Multi-Relational Data Mining.