How we’re helping preserve the genetic information of endangered species with AI

Lizzie Dorfman

Group Product Manager

Andrew Carroll

Product Lead, Genomics

Scientists predict that one million species may be at risk of extinction. If we don’t accurately capture their genetic information now, we risk losing many of them forever and destabilizing the ecosystems that provide us with food security, climate regulation and the biological foundations for modern medicine. Still, sequencing the genomes of millions of species on our planet is an extraordinarily difficult and expensive task — if not impossible —and it’s one the Vertebrate Genomes Project, led by Erich Jarvis at The Rockefeller University, is tackling with Google’s AI.

To help them and the Earth BioGenome Project achieve their goal of sequencing all known species on Earth, we’ve supplied funding, support and cutting-edge AI tools to help preserve the genetic code of 13 new endangered species across several animal classes, including mammals, birds, amphibians and reptiles — and recently announced more support. Here’s a look at nine of the species, whose genomes are free and available for use by conservation researchers.

Understanding a species’ genome could be the key to their conservation

By sequencing these animals’ genomes — their complete set of genetic instructions (DNA) — scientists are able to support efforts to preserve the genetic diversity of these animals and prevent further loss of their species. Researchers at the University of Otago, for instance, analyzed the genome of every living kākāpō, the world's only nocturnal and flightless parrot, and executed a breeding and conservation plan that is now successfully pulling it back from the brink of extinction. Genomes not only help us understand how a particular species has adapted to life in its environment; when we compare the DNA sequences of different species, biologists can also better understand the history of life on this planet. These insights are critical to future breakthroughs in conservation, agriculture, and even global health and disease prevention.

Understanding the biological picture of life

For more than a decade, Google researchers have been building AI technology to help scientists achieve what once seemed impossible, from forecasting floods to mapping the connections in the human brain. It took thirteen years and $3 billion to sequence the very first human genome. Now, we can sequence humans, animals and plants in a matter of days, with a few thousand dollars and with stellar accuracy. AI tools developed by Google, like DeepPolisher, DeepVariant and DeepConsensus played an important role in these improvements, and what once felt like a moonshot is within reach: a comprehensive catalogue that encapsulates a complete biological picture of life, changing science as we know it.

The technologies we build at Google have the strongest impact when scientists use them to solve real-world problems. To further the important mission of the Vertebrate Genome Lab and Earth BioGenome Project, Google.org recently named The Rockefeller University as a recipient of the AI for Science fund to expand the above number to 150 additional species — all to be openly released to the scientific community and public.