The gene known as T-urf13 has become a favorite of ID critics. It's supposed to be a de novo mitochondrial gene that rapidly evolved via random mutation from parts of a genome that coded for rRNA and also non-coding genes. And this would provide a strong challenge to Behe's claim that the evolution of de novo genes by random mutation is exceedingly rare. Perhaps this is the case. However, I've been doing a little reading, and it appears that the story might be a little more complicated. Consider some of the information from Higher Plant Mitochondria :
Moreover, within given plant families, it has been persuasively demonstrated that individual plant species likely represent evolutionary intermediates in an ongoing process of gene transfer from the mitochondrion to the nucleus (Brennicke et al. 1993 ; Gray 1995 ). Such gene transfer apparently occurs via RNA intermediates, presumably a vestige of earlier endosymbiotic processes. If this is the case, how would a newly introduced nuclear form of a gene then derive a means of transferring its product back to the mitochondrion? Analysis of "recently" transferred mitochondrial genes within the nucleus of rice has established the integration of introduced genes at duplicated sites already encoding mitochondrial proteins to allow, essentially, the requisitioning of the previous transit sequence (Kadowaki et al. 1996 ).
Is it possible that the genome that supposedly evolved into the T-urf13 gene was just an incomplete version of it? Is it possible that the rest of the gene resided in the nucleus and was "requisitioned" by the mitochondrion?