Pharmacogenomics studies the contribution of interindividual variations in DNA sequence to different drug responses (especially adverse drug reactions). Records of millions of genomic variations, mostly known as Single Nucleotide Polymorphisms (SNP), are available today in various overlapping and heterogeneous databases including private sources. KDD (knowledge discovery in databases) process is an attractive methodology for addressing the pharmacogenomic challenge. However, the complexity of this field makes it necessary to guide the KDD process by domain knowledge. Formal representation of domain knowledge in an ontology is proposed here as a means for guiding the data selection step and for semantically integrating data about genomic variations. The designed SNP-Ontology is used to build a SNP-dedicated knowledge base which integrates data on genomic variations whatever their original representations. A specific wrapper, the SNP-Converter, has been developed for managing of representation heterogeneity and for populating the knowledge base.