Google DeepMind maps 9 billion human DNA mutations in major genetics breakthrough

Google DeepMind officially introduced AlphaGenome Atlas model- a massive genomic database that provides functional predictions for every possible single nucleotide variant in the human genome.

The new AI model pre-calculate the compliance effect of all 9 billion single-letter genetic changes resulting in a substantial 1-petabyte dataset with ability to help scientists immediately query vast information.

The primary motive behind the new model is to help researchers easily use an integrative scoring framework for both coding and non-coding regions, allowing researchers to fast-track the high-value opportunities for research without sifting through thousands of data points.

Laura Covill and her team at the Broad Institute used the AVI score to candidate variant selection for unsolved rare disease research. The tool underlined a critical variant in the DNM1 gene predicting cryptic splice site activation.

Genomic lead at Google DeepMind Ziga Avsec said: “ If somebody is studying a disease, and they don’t have any idea about what cell types to look for or what molecular processes are impacted, then starting with an AVI score, which is a single score, is a great starting point to help you prioritize variants and try to find that needle in the haystack.”

Nonetheless, Atlas is available today through a website portal that requires zero coding skills, globalizing access for clinical researchers and medical scientists as part of sustained commitment to advancing genomic discovery.