These studies reported that relocation from the X chromosome to any autosomal arm, [X → A], is significantly more frequent than expected by chance, whereas any other relocation pattern ([A i → A j] and [A → X]), is significantly underrepresented.
We also looked for changes in gene expression and potential associations with the relocation pattern in other Drosophila species.
As described above for D. melanogaster, we looked for changes in the expression pattern between parental and duplicated genes associated with their relocation pattern in other Drosophila species.
Previous studies reported that the relocation pattern (i.e., change in location between parental and duplicated gene) of duplicate genes exhibits a nonrandom distribution in Drosophila (Betrán et al. 2002; Bai et al. 2007; Meisel et al. 2009; Vibranovski, Zhang, and Long 2009).
To test whether departures from the expected frequencies [X → A], [A → X], and [A i → A j] exist for the original relocation pattern, we first analyzed the chromosome distribution of the single copy and parental genes altogether.
We analyzed the relocation pattern in each Drosophila species genome to infer the ancestral gene locations and original relocation pattern (see Materials and Methods and supplementary table 1, Supplementary Material online).
Once we assigned the relocation pattern for the orthologous genes in every species, we further established the original relocation pattern for a gene family (supplementary table 2, Supplementary Material online), which was inferred by applying the maximum parsimony principle to the relocation patterns of the single species.
Furthermore, relocated genes achieve similar and extremely high (see previous section) levels of expression regardless of the relocation pattern (supplementary fig. 2, Supplementary Material online).
Briefly, we computed the probability of each retrotransposition or relocation pattern under the null hypothesis that all insertions occur randomly via simulation.
All of these studies concluded that selective forces might be operating across the whole Drosophila genome, which would favor the [X → A] relocation pattern (see table 1).
