Enabling efficient analysis of biobank-scale data with genotype representation graphs (Data column)

I came across the ACM newsletter here. (Association of Computing Machinery) The news, which perfectly demonstrates that Big data isn't big anymore.:) In fact, there was always a lot of data in genetics after scientists learned how to decipher the genome. But this data had to be able to store somewhere and quickly shovel in order to extract insights. And now researchers at Cornell have developed a new data compression method that allows large genomic datasets to be stored on local computers. This data used to weigh hundreds of terabytes, but now it's compressed into gigabytes. This method is called the Genotype Representation Graph. (GRG)described in articlepublished 5 December 2024 Year in Nature Computational Science. GRG uses graphs to represent genotypes, which allows for compact and intuitive encoding of genome information, as well as performing calculations without the need for data compression. This makes the analysis of biobank data more efficient and less costly. Unlike traditional matrix representations, GRG captures connections between individuals through common mutations in their genomes. The method was developed to solve the problem of data volume growth, which can now reach petabytes due to the increasing availability of whole genome sequences. GRG provides scalability and accurate data representation, enabling complex analyses that were previously unavailable due to high computational cost.

The research has already caught the attention of the scientific community, and other scientists have begun testing the method on different data sets. The work is supported by a grant from the US National Institutes of Health.

#Data #Math #Software #Whitepaper