Two dimensional sequencing and machine learning to maximise genetic marker detection

Principal Investigator

Prof. Brett Lidbury


The John Curtin School of Medical Research, Australian National University, Canberra, Australia

Start date

July 2016

Background and aim

Thanks to recent technological advances,  the genome of any individual (a complete set of their DNA) can be rapidly surveyed from a single sample of tissue using machine automation. This technology has  allowed genome-wide studies to be undertaken in many different diseases (read a review). Human genomes vary greatly, however, so hundreds or thousands of patients with a particular disease and healthy controls need to be recruited for valid results, and sequencing studies can be very expensive indeed. This is a particular problem in illnesses like ME/CFS where funding is scarce and where it can be hard to access large numbers of well-defined and clinically assessed patients. One way to avoid recruiting thousands of research participants, reducing costs and time, is DNA pooling. The basic idea is that DNA from many different patients is pooled together into a single DNA mixture, which is then sequenced, reducing considerably the ‘library set-up’ costs involved in preparing samples individually.

Prof Brett Lidbury and colleagues at the John Curtin School of Medical Research, Australian National University in Canberra are conducting an ongoing programme, funded by charities in Australia, which aims to find biomarkers for ME/CFS using a range of sources – bioinformatics, genetics and pathological testing. Prof Lidbury has a background in virus–host interactions, particularly the Ross River virus which is endogenous to Australia and can cause long term post-viral syndromes, and has been suspected to have a role in ME/CFS. Today, much of his work is on ME/CFS, and his team benefits from access to a large well-defined group of Canadian Consensus Criteria-defined patients recruited by the CFS Discovery Clinic in Victoria, for whom full clinical histories, and pathological and physiological data are available. In fact, it recently published a report on the co-occurrence of postural orthostatic tachycardia syndrome and ME/CFS using data from this cohort (read more).

ME Research UK has provided funding to the group to acquire additional genetic data using DNA pooling. They will apply two dimensional DNA sequencing to attempt to identify ME/CFS-associated genetic changes across the entire genome in a clinically well-defined group of 100 patients and up to 40 controls. Two DNA pooling techniques will be used – a pooling/bootstrap genome-wide method which has been developed and used for the detection of genetic markers of Alzheimer’s Disease by Prof Lidbury’s collaborators on this project, Dr Mastronardi and Prof Arcos-Burgos of the Genomics and Predictive Medicine Group at the John Curtin School (read more), and a 2D DNA pooling method for rare variant detection. The team recently completed a pilot study which successfully applied machine learning methods to pathology and clinical data alone, revealing patterns associated both with the presence and severity of ME/CFS. So, when the genetic data is available from the DNA pooling studies, it will be possible to apply similar machine learning techniques and statistical analyses to an integrated data set combining genetic, clinical and pathological information.

The team at the Australian National University: (L to R) Dr Alice Richardson, Dr Claudio Mastronardi, Dr Guifang Shang, Prof Mauricio Arcos-Burgos, Prof Brett Lidbury and Ben Signor
The team at the Australian National University: (L to R) Dr Alice Richardson, Dr Claudio Mastronardi, Dr Guifang Shang, Prof Mauricio Arcos-Burgos, Prof Brett Lidbury and Ben Signor

As Prof Lidbury explained in a recent interview, “We’re looking for patterns in the data which can help with biomarker pattern detection and provide clues to disease mechanisms. With complex diseases like this, we need to look at many factors simultaneously, and cross-disciplinary studies are necessary to integrate the findings from different avenues of research investigation. I’ve been very fortunate to have spent a large proportion of my career involved in curiosity-driven fundamental research. My plan now is to work on tangible solutions that translate to patient benefit. For the moment this is focused on biomarker discovery, which may (hopefully) also lead to ideas for therapeutic interventions.”

Further reading
Meet the researchers: Dr Brett Lidbury. ME Australia, 2016 (read more)
Systems genetics approaches to understand complex traits. Nature Reviews: genetics, 2014 (read more)
Vélez, J, et al. Pooling/bootstrap-based GWAS (pbGWAS) identifies new loci modifying the age of onset in PSEN1 p. Glu280Ala Alzheimer’s disease. Molecular psychiatry, 2013 (read more)
Reynolds, G, et al. Comorbidity of postural orthostatic tachycardia syndrome and chronic fatigue syndrome in an Australian cohort. Journal of Internal Medicine, 2014 (read more).

Verified by MonsterInsights