Is the guinea-pig a rodent?

The guinea-pig (Cavia porcellus), traditionally classified as a New World hystricomorph rodent, often shows anomalous morphological and molecular features in comparison with other eutherian mammals. For example, its insulin differs from that of other mammals in anabolic and growth-promoting activities and in its capability to form hexamers. Indeed, the literature about the molecular evolution of guinea-pigs abounds in references to 'convergent evolution', 'extremely rapid rates of substitution', and 'unique evolutionary mechanisms'. These claims are based on the assumption that the guinea-pig is a rodent. Our phylogenetic analyses of amino-acid sequence data, however, imply that the guinea-pig diverged before the separation of the primates and the artiodactyls from the myomorph rodents (rats and mice). If true, then the myomorphs and the caviomorphs do not constitute a natural clade, and the Caviomorpha (or the Histricomorpha) should be elevated in taxonomical rank and regarded as a separate mammalian order distinct from the Rodentia. If, as suggested by recent data, the myomorphs branched off before the divergence among the carnivores, lagomorphs, artiodactyls and primates, then the new order would represent an early divergence in eutherian radiation.