A second, and probably related, effect is that proteins encoded by the human genome tend to contain more multiple domains than do those encoded by C. elegans. This is especially true for transmembrane proteins with large extracellular regions, which are presumably involved in cell-cell interactions of one kind or another. Such classes of nematode proteins may be generally simpler and less multifunctional than their mammalian counterparts, and therefore a greater variety would need to be encoded by separate genes.
A third consideration arises from the apparent extensive duplication of genes in the C. elegans genome. Many genes and small genetic regions appear to have experienced piecemeal duplication during the evolution of this species [2]. The level of duplication appears to be higher in the autosomal chromosome-arm regions of the genome, which are also regions where genes encoding conserved eukaryotic core functions are underrepresented [2,19]. Much of the duplication seems to be of ancient origin, which implies that selective forces maintain the present high number of gene pairs, whatever their origins may be.