And They Names it Evo — an AI trained on trillions of bases

  • Thread starter Thread starter jedishrfu
  • Start date Start date
Click For Summary

SUMMARY

Evo is an open-source AI system developed in late 2025, trained on trillions of bacterial genome bases to predict gene sequences and suggest novel proteins within bacterial clusters. Its effectiveness relies on the genomic clustering characteristic of bacteria, which does not extend to complex eukaryotic genomes with intricate structures. The AI demonstrates advanced capabilities in genomic sequence prediction specifically for prokaryotic organisms. The discussion highlights the global health impact of bacterial infections, responsible for 7.7 million deaths annually, emphasizing the dual role of bacteria as both pathogens and essential contributors to human health.

PREREQUISITES

  • Genomic sequence analysis of bacterial genomes
  • Understanding of AI models trained on biological data
  • Knowledge of prokaryotic gene clustering and operon structures
  • Familiarity with protein prediction algorithms

NEXT STEPS

  • Explore Evo’s open-source codebase and training datasets
  • Research AI applications in complex eukaryotic genome analysis
  • Study bacterial operon prediction and gene cluster identification techniques
  • Investigate protein structure prediction tools integrated with genomic AI models

USEFUL FOR

Computational biologists, bioinformaticians, AI researchers in genomics, and microbiologists focused on bacterial genetics and protein discovery will benefit from this discussion. It is also valuable for healthcare professionals interested in the implications of bacterial genomics on infectious disease management and antibiotic development.

jedishrfu
Mentor
Insights Author
Messages
15,753
Reaction score
10,639
https://arstechnica.com/science/202...open-source-ai-trained-on-trillions-of-bases/

This is for @Evo: Hello Mrs Calabash wherever you.

—- Ars Technica article exerpt follows

Late in 2025, we covered the development of an AI system called Evo that was trained on massive numbers of bacterial genomes. So many that, when prompted with sequences from a cluster of related genes, it could correctly identify the next one or suggest a completely novel protein.

That system worked because bacteria tend to cluster related genes together—something that’s not true in organisms with complex cells, which tend to have equally complex genome structures. Given that, our coverage noted, “It’s not clear that this approach will work with more complex genomes.”
 
Reply
  • Like
  • Informative
Likes   Reactions: Astronuc, pinball1970, BillTre and 2 others
Biology news on Phys.org
Wow, my thread is front page news! Now if anyone here would like to discuss it to keep it there.

We're shooting for like a couple of weeks. :-)
 
jedishrfu said:
Wow, my thread is front page news! Now if anyone here would like to discuss it to keep it there.

We're shooting for like a couple of weeks. :-)
Well, as we used to say in the olden says; "How does this knowledge affect the price of rice in China?"
As an oldster, I'm worried about things trying to kill me.

According to this article in the Lancet;

"Each year, an estimated 7.7 million deaths globally are caused by bacterial infections—one in eight of all global deaths—making bacterial infections the second largest cause of death globally."
{bolding mine}

Now, if I were not living the high life here in the naughty colony, I'd be worried about such a thing. But I do, so I don't.
 
On the other hand, according to this article, we'd all be dead without bacteria.

"Some bacteria are indisputably bad. But others boost our immunity, protect us from infection and produce the enzymes we need to digest our food. Without these bacteria, we wouldn't survive, says Dr. Martin Blaser, author of the new book "Missing Microbes.""
 
Death will come for each of us, with or without bacteria or other pathogens.
Screenshot 2026-03-07 at 3.52.57 PM.webp
 

Similar threads

Synthesis of a re-designed E. coli genome
  • · Replies 3 ·
Replies
3
Views
3K
Interesting Path to Eukaryotic Cell Complexity Proposed
  • · Replies 4 ·
Replies
4
Views
3K