---
slug: "deepmind-alphagenome-2-0456dde7"
title: "Google DeepMind releases AlphaGenome 2 gene sequence model: nearly doubles the ability to identify pathogenic variants"
headline: "Google DeepMind releases AlphaGenome 2 gene sequence model: nearly doubles the ability to identify pathogenic variants"
canonical: "https://silinex.xyz/wire/deepmind-alphagenome-2-0456dde7"
date_published: "2026-11-11T15:19:51.000+00:00"
date_modified: "2026-11-11T16:21:26.422+00:00"
author: "IT之家"
publisher: "Silicon Nexus"
category: "ai"
summary: "Compiled from publicly available reports such as IT Home: Google DeepMind has released a new generation DNA sequence model, AlphaGenome 2, which can predict the effects of DNA sequences and their variations on specifi…"
key_facts:
  - "IT Home, November 11 – Google DeepMind has released a new generation DNA sequence model, AlphaGenome 2, which can predict the effects of DNA sequences and their variations on specific human cell types, providing researchers with a new tool to explore the cellular drivers of diseases."
  - "Every cell in the human body carries the same DNA blueprint, but different cell types perform specific functions by switching different genes on and off."
  - "The Human Cell Atlas project worldwide has mapped the gene expression of millions of cells, but predicting how subtle DNA variations affect cellular gene expression remains an unsolved challenge."
  - "Previously, Google had launched the first-generation AlphaGenome, which can predict the impact of DNA variants on the molecular activity of broad tissues, and released the AlphaGenome Atlas, covering predictions for all 9 billion single-letter variants in the human genome, aiding research on rare diseases, complex traits, and basic biology."
---
# Google DeepMind releases AlphaGenome 2 gene sequence model: nearly doubles the ability to identify pathogenic variants

## Silicon TL;DR

1. IT Home, November 11 – Google DeepMind has released a new generation DNA sequence model, AlphaGenome 2, which can predict the effects of DNA sequences and their variations on specific human cell types, providing researchers with a new tool to explore the cellular drivers of diseases.
2. Every cell in the human body carries the same DNA blueprint, but different cell types perform specific functions by switching different genes on and off.
3. The Human Cell Atlas project worldwide has mapped the gene expression of millions of cells, but predicting how subtle DNA variations affect cellular gene expression remains an unsolved challenge.
4. Previously, Google had launched the first-generation AlphaGenome, which can predict the impact of DNA variants on the molecular activity of broad tissues, and released the AlphaGenome Atlas, covering predictions for all 9 billion single-letter variants in the human genome, aiding research on rare diseases, complex traits, and basic biology.

## Wire brief

Compiled from publicly available reports such as IT Home: Google DeepMind has released a new generation DNA sequence model, AlphaGenome 2, which can predict the effects of DNA sequences and their variations on specific human cell types, providing researchers with a new tool to explore the cellular drivers of diseases.

In terms of the context of the events, every human cell carries the same DNA blueprint, but different cell types perform specific functions by switching different genes on and off. The global Human Cell Atlas project has mapped the gene expression of millions of cells, yet predicting how subtle DNA variations affect cellular gene expression remains an unsolved challenge. Previously, Google launched the first-generation AlphaGenome, which can predict the effects of DNA variations on molecular activity in broad tissues, and introduced the AlphaGenome Atlas, which covers predictions for all 9 billion single-letter variations in the human genome, aiding research in rare diseases, complex traits, and basic biology.

Public information also mentions: AlphaGenome 2 reduces from the average level of broad tissues to individual cell types and states, and reconstructs the model architecture. Its variant prediction performance achieves a stepwise improvement, capable of capturing variant effects in disease-relevant cells missed by broad tissue models, and can also make predictions for cell and disease states that the model has not seen. According to IT Home, compared with the previously optimal model, AlphaGenome 2 can identify nearly twice as many variants that can alter gene expression within cells. This model is trained on over 100 million single-cell maps from 68 types of tissues, transforming static cell maps into a predictive engine for human genetics. Currently, the model is available for non-commercial research through the AlphaGenome API, while commercial research is in private preview.

Subsequent observation points include: The first-generation AlphaGenome can read DNA sequences of millions of letters at once and simultaneously predict relevant properties across multiple tissue types and thousands of experiments. This parallel approach is relatively efficient but disperses the model's computational capacity, making it difficult to capture subtle differences between closely related cell types. Google has fundamentally restructured the model's logic for handling cell types. Silicon News Wire only provides a summary rewrite; details should still be taken from the original text.

## Source & notice

This Silicon News Wire brief paraphrases public reporting from **IT之家** for carbon readers and silicon citation. **It is not a verbatim full-text reprint.**

Read the original: [Google DeepMind releases AlphaGenome 2 gene sequence model: nearly doubles the ability to identify pathogenic variants](https://www.ithome.com/1/025/822.htm)