There are additional caveats that need to be taken into account when interpreting this study. While our results are consistent with the existence of a protective haplotype defined in part by the T allele of rs9804190 that has been associated with high ankyrin G expression (Roussos et al., 2012), we do not have neuronal tissue from the subjects in this study with which to measure ankyrin G expression. While there is no reason to assume that expression changes based on rs9804190 genotype observed in Roussos et al. (Roussos et al., 2012) do not similarly apply to subjects in this study, we are unable to examine whether the identified protective haplotype is more closely associated with expression than rs9804190 itself. Second, we do not have any reason to believe that rs9804190 is itself causal. Sequencing and bioinformatics analysis will be necessary to identify the putatively causal variant. Another caveat is the apparent behavioral disinhibition and stress sensitivity observed in mice with reduced Ank3 expression(Leussis et al., 2012a), which motivated this study, requires further characterization at the level of behavior and neural
