Scientists at Columbia University have created a new AI model—General Expression Transformer, or GET. This cutting-edge algorithm is poised to change the way we understand gene behaviour, regulation, and their vital roles in both health and disease.
While we’re familiar with ChatGPT (Generative Pre-trained Transformer) in the realm of language processing, GET operates in the intricate world of genes. Just as GPT understands the complexities of human language, GET decodes the “grammar” of genes, revealing how they are activated or deactivated, ultimately influencing the proteins our bodies produce to fuel every action we perform.
Proteins, far from being simple building blocks, are incredibly complex structures with 3D shapes. Understanding these structures is vital for advancements in medicine, and AI tools like AlphaFold2 (and its successor, AlphaFold3) have already paved the way for predicting protein structures. Though still in its early stages, GET has the potential to rival such tools by predicting gene regulation in unprecedented detail.
What sets GET apart is its ability to predict gene behaviour in the context of genetic diseases like cancer. The model has been trained on an extensive dataset of 1.3 million human cells from 213 variants, equipping it with the capacity to predict gene regulation across previously untrained cell types. This could revolutionize gene therapies, allowing scientists to target specific cell variants with precision—without affecting surrounding healthy cells.
The possibilities are endless: with GET, researchers can prioritise genes implicated in diseases, speeding up the identification of genetic mutations and fueling accelerated progress in disease treatment. The model’s potential to drive gene therapy research is truly transformative, and the field is bracing for its impact on the future of personalized medicine and genetic disease management.
As the GET model continues to evolve, it could be a game-changer for gene therapies, offering a new frontier in our fight against genetic disorders and cancers, and paving the way for a future where precise, targeted treatments are the norm.