Google DeepMind has launched AlphaGenome, an advanced artificial intelligence model designed to predict the effects of single nucleotide variants and mutations across human DNA sequences. The model processes up to one million base pairs at once, enabling comprehensive analysis of both protein-coding and non-coding regions of the genome, including the previously elusive "dark matter" or regulatory genome. AlphaGenome predicts thousands of molecular properties related to gene regulation, RNA splicing, and disease risk, offering insights into gene expression and the impact of genetic changes on biological processes. Available via an API for non-commercial research, AlphaGenome aims to accelerate biological discoveries, synthetic biology, and the identification of cancer-related genes. The model combines convolutional layers with transformer architectures to deliver high-resolution, large-scale DNA sequence analysis, outperforming or matching existing genomics tools. This development represents a significant step toward AI-driven precision medicine and could transform disease-risk prediction and genomic research worldwide.
ApexOracle integrates three foundational representation modules. The genomic encoder employs Evo2, a DNA language model pretrained on genomes spanning all domains of life, to transform a pathogen's entire genome into a numerical representation that captures genotypic hallmarks.
AI Uncovers New Alzheimer's Genes In a groundbreaking development, a machine learning scan has identified novel risk loci for Alzheimer's disease within a massive Genome-Wide Association Study (GWAS). Could precision medicine accelerate the fight against memory loss? #AI
NIST Releases Trove of Genetic Data to Spur Cancer Research https://t.co/hxh8hqnnSo