Genomics research, biological samples and DNA

An important unresolved question in autism research is whether genomic information can assist in early and accurate diagnosis of autism. The objectives of this strategic project, commissioned in late 2015, were to receipt and process biospecimens for the Australian Autism Biobank, derive and evaluate systems genomics-based predictors for autism using Biobank samples, and to integrate these with a clinical diagnostic instrument based on behavioural surveillance.

The primary research implication from this project is that, as anticipated, much larger genomic datasets will be needed to fully evaluate genomics predictors in autism. Achieving this goal will require coordinated international collaboration to achieve data harmonisation and the necessary sample sizes.

A number of potentially important clinical implications emerged from our analysis of gut microbiome data in the Biobank, which supports a 'top-down' model of causality, whereby restricted diet in autism influences the gut microbiome and stool consistency, contrary to evidence from animal models suggesting a causal effect (i.e. 'bottom-up') of the microbiome on autism-related behaviours.

In other work we looked at interactions between the lipidome and genetic and environmental factors in autism. We found that various plasma lipids were associated with autism diagnosis, IQ and sleep disturbances. Sleep problems had the strongest lipid associations, whereas autism had weaker associations with lipids. Notably, lower levels of some long-chain polyunsaturated fatty acids appeared to contribute to sleep problems.