What’s hiding in 98% of our DNA?

DeepMind’s AlphaGenome is shining new light on the genome’s “dark matter” with powerful AI insights.

In partnership with

Transform your hiring with Flipped.ai – the hiring Co-Pilot that's 100X faster. Automate hiring, from job posts to candidate matches, using our Generative AI platform. Get your free Hiring Co-Pilot.

Dear Reader,

What if unlocking the mysteries of rare diseases or designing next-generation therapies was as simple as asking an AI?

Flipped.ai’s weekly newsletter is read by more than 75,000 professionals, entrepreneurs, decision-makers, and investors around the world.

This week, we’re diving into DeepMind’s AlphaGenome—a breakthrough AI model that decodes the “dark matter” of our DNA, unlocking how gene regulation works and opening doors to faster disease research, precision medicine, and bioengineering innovations.

Create How-to Videos in Seconds with AI

Stop wasting time on repetitive explanations. Guidde’s AI creates stunning video guides in seconds—11x faster.

  • Turn boring docs into visual masterpieces

  • Save hours with AI-powered automation

  • Share or embed your guide anywhere

How it works: Click capture on the browser extension, and Guidde auto-generates step-by-step video guides with visuals, voiceover, and a call to action.

What’s AlphaGenome all about?

Imagine your DNA as a massive instruction manual for life. Only 2% of it tells your cells how to build proteins (AlphaFold’s playground), but the other 98%? It’s like the "dark matter" of your genome, controlling how and when genes do their thing. Small changes in this non-coding DNA can flip switches that lead to diseases like cancer or rare genetic disorders. But decoding it? That’s been a puzzle scientists have been scratching their heads over forever.

Enter AlphaGenome, DeepMind’s shiny new AI that predicts how tiny DNA changes (called variants or mutations) mess with gene regulation. It’s like giving scientists a superpower to zoom into the genome at base-pair level and see what’s going on—across a million DNA letters at once!

Google DeepMind

“This is one of the most fundamental problems not just in biology—in all of science.”
— Pushmeet Kohli, Head of AI for Science, DeepMind

Why AlphaGenome is a big deal?

AlphaGenome isn’t just another AI model; it’s a unifying force in genomics. Here’s why it’s got researchers buzzing:

1. Long-Range DNA Vision with Pinpoint Accuracy: AlphaGenome can analyze up to 1 million base pairs of DNA (that’s a lot of genetic code) and make predictions down to individual letters. Older models had to pick between seeing the big picture or zooming in—they couldn’t do both. AlphaGenome does, and it does it fast, using half the computing power of its predecessor, Enformer. Talk about efficiency!

2. A one-stop shop for gene regulation: This AI doesn’t just predict one thing—it covers thousands of molecular properties, like:

i) Where genes start and stop in different cells.

ii) How much RNA is produced.

iii) Which DNA bits are accessible or bound by proteins.

iv) How DNA folds in 3D to interact with itself.

It’s like having a Swiss Army knife for genomics research. Scientists can now explore multiple angles of a DNA variant with a single tool, saving time and brainpower.

3. Splice junctions? handled!: For the first time, AlphaGenome can predict splice junctions—the spots where RNA gets cut and pasted during gene processing. Mess-ups here can cause serious diseases like spinal muscular atrophy or cystic fibrosis. AlphaGenome not only spots these junctions but also predicts how mutations might throw them off. That’s a huge leap for understanding rare genetic diseases.

4. Outperforming the competition: AlphaGenome went head-to-head with other top models and crushed it, beating them in 24 out of 26 benchmarks. Whether it’s predicting gene expression, DNA folding, or splicing changes, this model is the new gold standard. And unlike specialized models, it does all of this in one go. Dr. Caleb Lareau from Memorial Sloan Kettering Cancer Center called it “a milestone for the field.” We’re not arguing!

Real-world impact: What AlphaGenome means for you

Okay, so AlphaGenome is a beast in the lab, but why should you care? Whether you’re an AI pro or just fascinated by the future of medicine, here’s how this tech could change the game:

Unlocking disease mysteries

By pinpointing how DNA variants disrupt gene regulation, AlphaGenome could help scientists find the root causes of diseases like cancer or rare genetic disorders. For example, it nailed the mechanism of a mutation linked to T-cell leukemia, showing how it flips on a cancer-causing gene. This could lead to new treatments or even ways to prevent diseases before they start.

Designing DNA like Lego

Ever heard of synthetic biology? AlphaGenome could help scientists design custom DNA sequences to turn genes on or off in specific cells—like switching on a healing gene in nerve cells but not muscle cells. This could open doors to next-level therapies or even bioengineered solutions for global problems.

Speeding up research

Google DeepMind: Animation showing AlphaGenome taking one million DNA letters as input and predicting diverse molecular properties across different tissues and cell types.

With AlphaGenome’s API now available for non-commercial research, labs worldwide can start probing DNA variants faster than ever. No need to juggle multiple models or wait weeks for results. This means quicker discoveries and more answers about how our genomes work.

“AlphaGenome will be a powerful tool for the field. Determining the relevance of non-coding variants can be extremely challenging. This tool will provide a crucial piece of the puzzle.”
— Professor Marc Mansour, University College London

Google DeepMind: Bar graph showing AlphaGenome’s relative improvements on selected DNA sequence and variant effect tasks, compared against results for the current best methods in each category.

The catch? It’s not perfect (yet).

No AI is flawless, and AlphaGenome has some limits. It struggles with super-distant DNA interactions (think 100,000+ letters away) and isn’t great at predicting cell-specific patterns yet. It’s also not ready for personal genome predictions—like telling you if a mutation means you’ll get sick. And while it can spot molecular changes, it doesn’t fully explain how those lead to complex diseases influenced by lifestyle or environment.

But DeepMind’s already working on these gaps, and with feedback from the research community, AlphaGenome’s only going to get better.

How can you get involved?

Excited about AlphaGenome? Here’s how you can jump in:

  • Researchers: The AlphaGenome API is open for non-commercial use. Head to DeepMind’s community forum to explore use cases, ask questions, or share feedback.

  • AI pros: This model’s hybrid architecture (convolutional layers + transformers) is a masterclass in handling massive sequences. Dig into the preprint for the techy details and get inspired for your own projects.

  • Curious minds: Follow DeepMind’s updates for the full model release and keep an eye on how AlphaGenome shapes the future of genomics.

The big picture

Twenty years after the human genome was sequenced, most of it remains a mystery. AlphaGenome is a flashlight into that darkness, helping decode how our DNA really works. From understanding rare diseases to designing life-saving therapies, this could be another AlphaFold-level breakthrough.

What do you think? Is AlphaGenome the future of genomics?
Hit reply; we’d love your take.

Stay curious.

Want to get your product in front of 75,000+ professionals, entrepreneurs, decision-makers, and investors around the world? 🚀

If you are interested in sponsoring, contact us at [email protected].

Thank you for being part of our community, and we look forward to continuing this journey of growth and innovation together!

Best regards,

Flipped.ai Editorial Team