Heteroplasmic sites in C. convexus
Coverage across the C. convexus mitochondrial genome was 450× on average (Figure 2). A total of four potential heteroplasmic sites were identified in C. convexus (Figure 2, Table 2). One site was an A/T polymorphism located at the boundary between A2 and T8 homopolymers (position 6998), which are known to be error-prone in 454 sequencing. Therefore, this putative heteroplasmy most likely reflects sequencing/mapping errors or artifacts.
Table 2  Mitochondrial sequence heteroplasmies in Cylisticus convexus
Position  Reference Allele  Alternative Allele  Alt. Frequency  Coverage  Notes
6998  T  A  0.25  526  nad2 substitution (F118I) located at boundary between A2 and T8 homopolymers
9309*  A  G  0.46  277  tRNA Leu1/Leu2 anticodon shared with T. rathkei
11755*  G  C  0.47  273  tRNA Arg/Gly anticodon shared with T. rathkei; also causes cox3 substitution (D274H)
12160*  G  A  0.47  806  tRNA Ala/Val anticodon shared with T. rathkei
Alt. frequency, frequency of the alternative allele in the sequence reads. *Heteroplasmic sites that were confirmed by Sanger sequencing. The other three heteroplasmic sites occurred in the anticodon of putative tRNA genes (positions 9309, 11755, and 12160), and each allele occurred at a frequency of approximately 0.5 in all three cases. Interestingly, they appear to be orthologous to the three confirmed heteroplasmic sites in T. rathkei tRNA anticodons. The heteroplasmy at position 12160 has previously been reported to be widespread among terrestrial isopods, including C. convexus (Doublet et al. 2008). The heteroplasmies at positions 9309 and 11755 were confirmed by Sanger sequencing with the original DNA pool used for 454 sequencing, as well as a set of 12 individual samples originating from diverse geographic locations (Figure S3).
The heteroplasmy at position 9309 is >2 kb away from the other heteroplasmies, which precluded identification of haplotype phase. By contrast, the heteroplasmies at positions 11755 and 12160 are ∼400 bp away from each other, which falls in the range of 454 sequencing read length, thereby enabling identification of haplotype phase. Consistently, we identified 35 reads covering both sites: 18 contained both reference alleles (11755G and 12160G) and 17 contained both alternate alleles (11755C and 12160A) (Figure S4).