Researchers at CSIRO, Australia’s national science agency, have made significant strides in understanding the genetic underpinnings of Alzheimer’s disease through the use of artificial intelligence (AI). By employing CSIRO’s VariantSpark and BitEpi tools, scientists at the Australian e-Health Center have identified two previously unknown genetic variants associated with Alzheimer’s disease, as well as 95 new gene interactions that potentially modulate the effects of these variants.
The identification of genetic variants plays a crucial role in predicting the occurrence, severity, and potential treatment options for neurodegenerative diseases like Alzheimer’s. However, it is important to note that these variants alone do not fully explain the heritability of Alzheimer’s and other related conditions. The interactions between different variants, known as epistasis, are believed to contribute to the onset and manifestation of the disease.
Traditionally, variants were measured based on their cumulative effect, examining how the combination of one gene with another increased the likelihood or expression of the disease. However, this approach did not account for the protective interactions between genes in relation to Alzheimer’s.
Dr. Natalie Twine, a CSIRO Research Scientist and senior author on the paper published in Scientific Reports, explained that certain gene interactions can actually protect against Alzheimer’s. She further emphasized that by utilizing BitEpi, the team was able to identify these interactions and shed light on previously unknown aspects of Alzheimer’s heritability.
Alzheimer’s disease is the most prevalent form of dementia, and it poses a significant health challenge. As of 2022, there were over 400,000 individuals living with dementia in Australia, and it is expected that these numbers will double by 2058 due to the aging population.
Dr. Mischa Lundberg, a CSIRO post-doctoral fellow and lead author on the paper, highlighted the significance of incorporating epistatic interactions in their research. By doing so, they were able to capture 10.41% more phenotypic variance compared to previous methods. This increase in capturing the drivers of the disease is crucial for Alzheimer’s research, as it enables the identification of individuals at risk at an earlier stage, allowing for timely intervention.
Moving forward, CSIRO plans to continue testing and applying VariantSpark and BitEpi tools to address existing challenges. One such challenge is the fragmented storage of genomic information in different geographical locations, limiting data sharing due to privacy concerns. This makes it difficult to consolidate data for more comprehensive research studies.
VariantSpark offers a solution called federated learning, where a machine learning model can be generated using data sources from different locations without the need to reveal the entire dataset. This approach will facilitate collaboration among researchers and enhance the power of research studies in the field of Alzheimer’s and other complex diseases.
The discovery of these two new genetic variants and the understanding of their interactions brings researchers one step closer to unraveling the mysteries of Alzheimer’s disease. With AI and innovative tools like VariantSpark and BitEpi, scientists have the potential to revolutionize disease prediction, treatment, and patient care in the future
1. Source: Coherent Market Insights, Public sources, Desk research
2. We have leveraged AI tools to mine information and compile it