--- title: AlphaGenome 2: Deciphering how cells read DNA to better... description: Google DeepMind has announced AlphaGenome 2, a next-generation DNA sequence model that predicts how genetic variants affect specific human cell types rather... canonical: https://daily.dev/posts/alphagenome-2-deciphering-how-cells-read-dna-to-better-understand-disease-rjek6qhhz twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: AlphaGenome 2: Deciphering how cells read DNA to better understand disease | daily.dev og:description: Google DeepMind has announced AlphaGenome 2, a next-generation DNA sequence model that predicts how genetic variants affect specific human cell types rather... og:url: https://daily.dev/posts/alphagenome-2-deciphering-how-cells-read-dna-to-better-understand-disease-rjek6qhhz og:image: https://api.daily.dev/og/posts/Rjek6QHHz.png og:image:alt: AlphaGenome 2: Deciphering how cells read DNA to better understand disease og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # AlphaGenome 2: Deciphering how cells read DNA to better understand disease **[DeepMind](https://daily.dev/sources/dm)** · 9 min read · 1 upvotes · 0 comments ## Summary Google DeepMind has announced AlphaGenome 2, a next-generation DNA sequence model that predicts how genetic variants affect specific human cell types rather than broad tissue averages. It identifies nearly twice as many variants that change gene expression compared to the previous AlphaGenome model. The model was trained on more than 100 million individual cell profiles across 68 tissues and adds DNA methylation as a new predicted readout. On the TenK10K dataset of 28 immune cell types, it found causal variants in 30% of cases versus 16% for the prior model. It is available now for non-commercial research via the AlphaGenome API, with a private preview for commercial research, and AlphaGenome 2 Skills are coming to Google Antigravity. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is AlphaGenome 2 and how is it different from the original AlphaGenome model? AlphaGenome 2 is DeepMind's next-generation DNA sequence model that predicts variant effects for specific human cell types rather than broad tissue averages. Unlike the original parallel-prediction architecture, it is a conditional model that takes a target tissue, cell type or state as input alongside the DNA sequence, letting it capture subtle differences between closely related cell types and even impute missing experimental data types like chromatin accessibility. _Researchers evaluating cell-type-specific variant tools can track genomics model releases like this on daily.dev._ ### How much better is AlphaGenome 2 at identifying causal genetic variants compared to prior models? On the TenK10K dataset covering 28 primary immune cell types, AlphaGenome 2 identified causal variants in 30% of cases versus 16% for the original AlphaGenome fine-tuned on single-cell data, and correctly predicted directional gene effects 43% of the time versus 23% previously, roughly doubling detection performance on both measures. _Anyone comparing genomic variant-calling tools can follow benchmark updates like this on daily.dev._ ### Can AlphaGenome 2 make predictions for cell types or states it was never trained on? Yes, it generalizes to entirely new single-cell experiments without retraining by compressing a single-cell RNA-seq snapshot from just a few dozen cells into a cell state representation used to condition the model. In one example, it distinguished vaccine-activated T cells from resting ones using unseen COVID-19 mRNA vaccine data and found a severity-linked variant whose effect was stronger only in the activated state. _Developers building on foundation models for biology can keep up with generalization capabilities like this via daily.dev._ ## Similar posts on daily.dev - [AlphaGenome Decodes DNA's Hidden Secrets](https://daily.dev/posts/alphagenome-decodes-dna-s-hidden-secrets-5ltqipgdj) · IEEE Spectrum · 0 upvotes · 0 comments - [Large genome model: Open source AI trained on trillions of bases](https://daily.dev/posts/large-genome-model-open-source-ai-trained-on-trillions-of-bases-wp1aoyqmr) · Ars Technica · 0 upvotes · 0 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#machine-learning](https://daily.dev/tags/machine-learning), [#biotech](https://daily.dev/tags/biotech) [View this post on daily.dev](https://daily.dev/posts/alphagenome-2-deciphering-how-cells-read-dna-to-better-understand-disease-rjek6qhhz) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"AlphaGenome 2: Deciphering how cells read DNA to better understand disease","url":"https://daily.dev/posts/alphagenome-2-deciphering-how-cells-read-dna-to-better-understand-disease-rjek6qhhz","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/alphagenome-2-deciphering-how-cells-read-dna-to-better-understand-disease-rjek6qhhz"},"datePublished":"2026-11-12T18:10:03.733Z","dateModified":"2026-11-12T18:55:46.783Z","description":"Google DeepMind has announced AlphaGenome 2, a next-generation DNA sequence model that predicts how genetic variants affect specific human cell types rather...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/db2bf18964c70817427e007cf90b01d7?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/db2bf18964c70817427e007cf90b01d7?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"DeepMind","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"DeepMind","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/dm","url":"https://daily.dev/sources/dm"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/alphagenome-2-deciphering-how-cells-read-dna-to-better-understand-disease-rjek6qhhz","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai,machine-learning,biotech","timeRequired":"PT9M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"DeepMind","item":"https://daily.dev/sources/dm"},{"@type":"ListItem","position":3,"name":"AlphaGenome 2: Deciphering how cells read DNA to better understand disease"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/alphagenome-2-deciphering-how-cells-read-dna-to-better-understand-disease-rjek6qhhz#faq","mainEntity":[{"@type":"Question","name":"What is AlphaGenome 2 and how is it different from the original AlphaGenome model?","acceptedAnswer":{"@type":"Answer","text":"AlphaGenome 2 is DeepMind's next-generation DNA sequence model that predicts variant effects for specific human cell types rather than broad tissue averages. Unlike the original parallel-prediction architecture, it is a conditional model that takes a target tissue, cell type or state as input alongside the DNA sequence, letting it capture subtle differences between closely related cell types and even impute missing experimental data types like chromatin accessibility. Researchers evaluating cell-type-specific variant tools can track genomics model releases like this on daily.dev."}},{"@type":"Question","name":"How much better is AlphaGenome 2 at identifying causal genetic variants compared to prior models?","acceptedAnswer":{"@type":"Answer","text":"On the TenK10K dataset covering 28 primary immune cell types, AlphaGenome 2 identified causal variants in 30% of cases versus 16% for the original AlphaGenome fine-tuned on single-cell data, and correctly predicted directional gene effects 43% of the time versus 23% previously, roughly doubling detection performance on both measures. Anyone comparing genomic variant-calling tools can follow benchmark updates like this on daily.dev."}},{"@type":"Question","name":"Can AlphaGenome 2 make predictions for cell types or states it was never trained on?","acceptedAnswer":{"@type":"Answer","text":"Yes, it generalizes to entirely new single-cell experiments without retraining by compressing a single-cell RNA-seq snapshot from just a few dozen cells into a cell state representation used to condition the model. In one example, it distinguished vaccine-activated T cells from resting ones using unseen COVID-19 mRNA vaccine data and found a severity-linked variant whose effect was stronger only in the activated state. Developers building on foundation models for biology can keep up with generalization capabilities like this via daily.dev."}}]} ```