• 10 years of genomics research at Google Oct 16, 2025 A look back at a decade of research in genomics at Google, from developing AI tools that accurately read the code of life, creating comprehensive reference genomes and working with partners on real-world challenges from biodiversity to healthcare. • General summary Google Research has spent the last 10 years developing AI tools to help geneticists study genomes. • These tools improve the speed and accuracy of reading and understanding the code of life. • Expect continued advancements in healthcare and biodiversity as Google Research continues to develop new tools. • Google Research has spent the last 10 years developing AI tools to help geneticists study genomes. • These tools improve the speed and accuracy of reading and understanding the code of life.
Article Summaries:
- Google marks a decade of applying deep‑learning techniques to genomics, highlighting a series of AI‑driven tools that have accelerated genome sequencing and analysis. Beginning in 2015, the company released DeepVariant in 2018, a variant‑calling tool that helped produce the first complete human genome. In 2022, DeepConsensus improved long‑read accuracy, boosting throughput, while DeepVariant was used to polish the NIH Telomere‑to‑Telomere reference genome. Subsequent projects included the first draft of a human pangenome (2023) and the DeepPolisher tool (2025) for higher‑quality assemblies across species. The latest DeepSomatic tool now offers more accurate cancer‑variant detection, underscoring Google’s ongoing role in advancing genomic research and healthcare.
Sources: