|Dr Sarah Kummerfeld||Dr Alexander Drew||Dr Andre Minoche|
|Dr Kishore Kumar||Dr Lisa Ewans|
To expand the clinical utility of whole-genome sequencing by developing algorithms to annotate pathogenic loci in noncoding regions of the genome, applied to a range of rare monogenic disorders of substantial unmet need.
Rare diseases are highly complex, often life-threatening and chronically debilitating. While these conditions are individually rare, they cumulatively affect around 1 in 12 Australians. Today, using whole genome sequencing and focussing our interpretation on the 2% of the genome that is protein-coding, we have improved the diagnostic rates for rare diseases to 60% on average. This still leaves a substantial fraction of patients that remain undiagnosed. One of the major challenges lies in understanding the link between rare diseases, and noncoding genetic variation, i.e. variation in the 98% of the genome that we currently ignore.
This team led by Dr Sarah Kummerfeld, are developing new tools and algorithms to interrogate the non-protein-coding genome and make new discoveries that can be translated to the clinic. This includes understanding the role of non-protein-coding regions in disease, especially splice and regulatory regions, as well as developing improved approaches to identify all sizes and types of genomic variation from the nuclear and mitochondrial genomes, and improved tools to annotated an understand this information (www.seave.bio). Their research is a combination of bioinformatic method development, applied to a number of rare conditions, including renal, cardiac, mitochondrial, epilepsy and movement disorders.
Disease-focused research activities
The Translational Genomics Informatics Team conducts research in the following disease-focused areas:
- Kidney disease
- Mitochondrial disease
- Parkinson's and movement disorders
- Mendelian disorders
- Epileptic encephalopathy
- Health economics of Intellectual Disability
- Facioscapulohumeral muscular dystrophy
Bioinformatic tools and resources
Comprehensive characterisation of the genome
The comprehensive characterisation of genetic variation present in a patient’s genome is essential to improve diagnostic rates. This requires the identification of single nucleotide variants, short insertions and deletions, structural and copy number variants, regions of homozygosity, repeat-length typing, phasing and haplotyping, retroviral insertions and transposon jumping, mitochondrial genetic variation and assessment of heteroplasmy, and an assessment of the impact of both coding, and noncoding genetic variation.
The Translational Genome Informatics team is working on improved methods for comprehensive genome characterisation.
Flexible, cloud based genomic analysis - refynr
Analysing WGS data at scale, reproducibly, and to a consistent clinical quality is a substantial computational challenge. To meet this challenge, we developed refynr, a cloud based genomic analysis platform, utilising the DNAnexus platform. refynr is a collection of >100 software modules, and a workflow generator, which has the flexibility to analyse any sized data, from targeted panel to WGS, and for both rare disease, and cancer.
Copy number and structural variants - ClinSV
Whole Genome Sequencing has the potential to comprehensively identify genetic variants of all sizes across the genome, from single base changes to entire chromosome changes.
Dr André Minoche leads the development of ClinSV, a dedicated Structural Variant and Copy Number Variant analysis pipeline. ClinSV has high sensitivity (>95%), reproducibility (>95%) and low false positive rate (<5%) to identify CNV and SV from short read WGS. It does this through careful integration of supporting evidence from multiple variant callers, and from cases vs 500 controls, with a visualisation framework.
Ben Lundie, Dr Mike Field, Dr Michael Buckley, Prof Leslie Burnett, Prof Andreas Zankl, and Dr Greg Peters.
KCCG’s variant filtration platform - Seave
Making sense of the millions of genetic variants present in a genome is extremely challenging. To address this challenge, Dr Vel Gayevskiy has led the development of Seave (www.seave.bio), a researcher- and clinician-friendly, web-based platform for filtering variants. It allows variants from WGS, exomes or targeted sequencing to be annotated, and filtered according to inheritance models, pathogenicity, rarity, and implication in known disease. By making it straightforward to curate gene lists, users can restrict their search to genetic variants to only those genes associated with the patient phenotype. Importantly, it stores short variants alongside CNV and SV, allowing integrated variant queries to be performed. Seave is being widely used for diagnosing patients with Mendelian disorders, rare disease gene discovery, genomics education, inherited cancer risk assessment, and somatic cancer genomics.
Dr Velimir Gayevskiy was the main developer of Seave, who is working closely with Dr Mark Cowley and clinical geneticists, medical specialists and genetic pathologists continue to develop Seave to solve the needs of both rare disease, and cancer research. Access to Seave is on a collaborative basis.
Regions of homozygosity - ROHmer
Consanguineous individuals have higher rates of genetic disorders, due to the inheritance of large, identical blocks of the genome. Ms Clare Puttick developed ROHmer to identify regions of homozygosity, also known as long contiguous stretches of homozygosity (LCSH), to both restrict the clinical analyses to regions more likely to be relevant to disease, and to estimate to extent of consanguinity in a genome. This research was undertaken with Dr Mark Cowley, Dr Tony Roscioli, Dr Kishore Kumar, and Ben Lundie.
Mitochondrial heteroplasmy - mity
Mitochondria are the energy producing organelles in cells, and they have their own small, circular, 16Kb genome. Each cell has multiple mitochondria, and each mitochondrion has multiple copies of the genome, giving rise to mitochondrial heteroplasmy, where a variant can be present in 0-100% of all mitochondrial genomes. Once the heteroplasmy of a pathogenic mutation increases above a clinical threshold in a given tissue, symptoms of a mitochondrial disease can appear. Ms Clare Puttick and Dr Mark Cowley have developed mity, a mitochondrial variant analysis pipeline, for sensitive characterisation of heteroplasmy from WGS data, from blood, or other tissues. They have developed this pipeline in collaboration with colleagues Dr Kishore Kumar, Dr Ryan Davis, Dr Lisa Riley, Prof Carolyn Sue, and Prof John Christodoulou.
Methods for temporal dynamics of transcriptome behavior
Cellular responses to stimuli are rapid and continuous. Despite this, the vast majority of investigations of transcriptional responses during developmental transitions typically use long interval time courses. This limits the available interpretive power. Dr Brian Gloss and Beth Signal are investigating coding and noncoding expression dynamics at unprecedented temporal resolution (6-hourly) in differentiating mouse embryonic stem cells to better understand molecular processes and genome organization. This team have been exploring complex and dynamic molecular events underlying mammalian differentiation that can only be observed though a temporally resolved time course, including: temporally ordered coding and noncoding transcriptional alterations including novel short lived and cycling patterns of gene expression, transcription factor and lncRNA driven network interactions, coordinately expressed genomic partitions and rapid alternative splicing events..
Applications of this work include investigating molecular pathways involved in resistance to chemotherapy. Ms Beth Signal and Dr Gloss, with A/Prof. Alex Swarbrick are leading a research project in breast cancer treatment. High frequency sampling enables identification of potentially transient changes that occur particularly early in chemoresistance that otherwise would be missed. Whole transcriptome analysis enables identification of additional events, such as alternative splicing, intron-retention, and their resultant impact on gene expression. This team are developing of tools and pipelines for processing temporal transcriptomic data.
Detection of splicing branchpoints
Single nucleotide polymorphisms in the intronic regions of genes are generally not considered clinically pathogenic. However, variants that disrupt the branchpoint motif (where splicing factors bind) can result in a faulty gene product. Ms Beth Signal is developing a method to identify branchpoints and determine whether changes will be pathogenic.
Product development and clinical applications
KCCG combines disease-focussed, and methods-development research to develop the analytical capability to diagnose new diseases, or new types of genomic variation (eg CNV and SV).
Neuronal Stem Cells Division (Renal diseases), John Shine
SV analysis of Schizophrenia
Neuroscience Division (Parkinsons & Neurodegeneration), Antony Cooper,
WGS of Skeletal Dysplasias
Translational Bone group, Andreas Zankl
Genomic determinants of bone mineral density
Osteoporosis & Translational Research group, John Eisman
Sydney Children’s Hospital Network
Elizabeth (Emma) Palmer, Rani Sachdev
Kolling Insitute of Medical Research
Carolyn Sue, Ryan Davis
Genetics of Learning Disabilities (GOLD)
Murdoch Children’s Research Institute
Victor Chang Cardiac Research Institute
Diane Fatkin, Sally Dunwoodie, Eleni Giannoulatou
ANZAC Institute and Molecular Medicine Lab, Concord Hospital
St Vincent’s Hospital
Tim Furlong, Kathy Wu