The Commonwealth Scientific and Industrial Research Organisation (CSIRO) has announced a team of researchers have processed one trillion points of genomic data in its fight to pinpoint the location of specific disease-causing genes in the human genome.
Using VariantSpark, CSIRO's artificial intelligence-based library for genomic data analysis, the researchers are looking for a deeper understanding of complex diseases by analysing large genomic datasets.
"Our VariantSpark platform can analyse traits, such as diseases or susceptibilities, and uncover which genes may jointly cause them," CSIRO Bioinformatics Group leader Dr Denis Bauer said. "This can provide valuable information about how the disease works on a molecular level, which can ultimately lead to better treatments."
She said VariantSpark is being used to help determine what genes might be linked to cardiovascular disease, motor neurone disease, dementia, and Alzheimer's disease.
The researchers analysed a synthetic dataset of 100,000 individuals, processing one trillion data points of genomic data, over ten million variants, and 100 thousand samples at once.
VariantSpark uses Amazon Web Services (AWS) Lambda, an on-demand serverless computing service and the CSIRO's genomic files are all located in an S3 datalake. It's available for download via the AWS Marketplace.
"Our research shows VariantSpark is the only method able to scale to ultra-high dimensional genomic data in a manageable time," Bauer said. "It was able to process this information in 15 hours while it would take the fastest competitors likely more than 100,000 years to process such a volume of data."
Over in Western Australia, scientists have been using Microsoft Azure high performance computing to perform high-resolution HLA genotyping at high throughput for unrelated haematopoietic stem cell transplantation donor recruitment.
PathWest Laboratory Medicine WA provides a critical HLA genotyping service to recruit unrelated donors to the Australian Bone Marrow Donor Registry, which is then linked to global registries.
When a patient needs a stem cell transplant, their HLA genotype is compared with all potential donors on the global registries. Having more enlisted local donors with high-resolution HLA-typing increases the chance of finding a best matched local donor for patients.
The team in Perth also developed a new technique to genetically characterise a blood sample from an organ donor.
"In some cases, the inability to identify the exact mismatches present between the donor and the recipient may increase the risk of rejection, which may outweigh the benefit of transplant, and therefore patients may be excluded from organ allocation," senior scientist in charge at PathWest's Department of Clinical Immunology Dr Dianne De Santis said.
"The ability to HLA type the deceased organ donor to the same typing resolution as the patient at the time of organ allocation and identify the exact mismatches, provides an opportunity for patients that previously may not have been considered suitable for organ transplantation to be considered."
Elsewhere, researchers from RMIT University have developed electronic artificial skin they say reacts to pain just like real skin, with better prosthetics, smarter robotics, and non-invasive alternatives to skin grafts on the horizon thanks to the innovation.
RMIT said the device mimics the body's near-instant feedback response and can react to painful sensations with the same speed that nerve signals travel to the brain.
"Skin is our body's largest sensory organ, with complex features designed to send rapid-fire warning signals when anything hurts," lead researcher Professor Madhu Bhaskaran said. "We're sensing things all the time through the skin but our pain response only kicks in at a certain point, like when we touch something too hot or too sharp. No electronic technologies have been able to realistically mimic that very human feeling of pain -- until now."
Bhaskaran said the pain-sensing prototype was a significant advance towards next-generation biomedical technologies and intelligent robotics.
The research was filed as a provisional patent combining three technologies previously pioneered and patented by the team: Stretchable electronics, temperature-reactive coatings, and brain-mimicking memory, which are electronic memory cells that imitate the way the brain uses long-term memory to recall and retain previous information.
- Max Kelsen using Google's Cloud to help beat cancer through machine learning
- Aussie biotech Microba using AI to research gut microbiome
- IBM tests the use of artificial intelligence for breast cancer screenings
- Victoria to use AI to detect heart attacks on emergency calls
- QUT researchers develop AI to improve accuracy around eye-testing