Isolation of human disease genes is a challenging process and can often only be achieved by labor-intensive positional cloning techniques. Fortunately, there are alternative strategies for isolation of peroxisome biogenesis disorder genes. The first, functional complementation, was established as a viable approach by Fujiki and colleagues, who identified PAF-1, the first known peroxisome biogenesis disorder gene. The second strategy, computer-based homology probing, relies on (1) the fact that peroxisome assembly has been conserved throughout the evolution of eukaryotes, (2) knowledge of the amino acid sequences of an increasing number of yeast peroxisome assembly (PAS) genes, and (3) the existence of sequence data from large numbers of human genes. The recent development of the expressed sequence tag (EST) database (dbEST) is fulfilling the last of these requirements. We have applied the homology probing strategy in the search for candidate genes for the peroxisome biogenesis disorders by routinely screening the database of ESTs for genes with significant sequence similarity to yeast PAS genes. The validity of this approach is demonstrated by its use in identifying PXR1 as the human orthologue of the Pichia pastoris PAS8 gene and PXAAA1 as a human homologue of the Pichia pastoris PAS5 gene. Furthermore, detailed analysis of PXR1 has revealed that mutations in this gene are responsible for complementation group 2 of the peroxisome biogenesis disorders. The demonstration that human homologues of yeast PAS genes exist and that mutations in these genes cause peroxisome biogenesis disorders demonstrates that yeast pas mutants are accurate and useful models for the analysis of these diseases.