From genome to phenome: Autistic spectrum disorder as an example

A complete understanding of understanding of behavioural disorders will involve connecting the dots between all of levels of analysis between genes and behaviour – from genome, transcriptome, epigenome, and proteome to the brain, mind and behaviour. In the case of autistic spectrum disorder (ASD), the behavioural level of analysis has been useful in showing not only high heritability but also genetic heterogeneity among the three components of the autistic triangle (social, non-social and communicative problems); progress towards identifying genes responsible for this high heritability has been slow, in part because it seems likely that there are many genes of small effect. The brain level of analysis can be useful in understanding ASD to the extent that it is related to the behavioural problems that we call ASD. However, I disagree with the notion that, as compared to behaviour, the brain is a simpler endophenotype, more heritable, or caused by a smaller number genes. Thus, it will not be easier to identify replicable gene associations at the brain level of analysis and if such genes are found the crucial question is whether the genes will be associated with ASD behaviours. It seems much more likely that genes will be identified for ASD at the behavioural level, and then these genes will be used as small footholds in the climb towards understanding the complexities of the brain as it relates to ASD.