← All news
·1 min read·deepmind

DeepMind announces AlphaFold 3: AI predicts interactions between proteins and other molecules

New version of the AI system predicts protein structures and interactions with DNA, RNA, and other molecules. Available free for academic researchers.

DeepMind launches AlphaFold 3 with molecular interaction prediction

Google DeepMind announced on May 8, 2024, the launch of AlphaFold 3, the latest version of its artificial intelligence system for protein structure prediction. Unlike previous versions, AlphaFold 3 can model not only proteins but also their interactions with DNA, RNA, and other molecules — a significant advance for computational biology and drug discovery.

What's new?

While the original AlphaFold (released in 2020) revolutionized protein structure prediction, AlphaFold 3 goes further by simulating how molecules bind to biological targets. This ability to predict molecular interactions is crucial for understanding cellular processes and developing new drugs.

Free access for academic research

The tool is freely available to academic researchers, who can access it via a server maintained by the lab. DeepMind hopes open access will accelerate discoveries in areas such as neglected diseases, structural biology, and protein engineering.

Impact on drug discovery

AlphaFold 3 can dramatically reduce the time needed to identify drug targets and test drug candidates in silico. The ability to simulate interactions between proteins and potential drugs allows researchers to prioritize the most promising compounds before costly and time-consuming lab tests.

Next steps

DeepMind intends to continue improving the model and expanding its database. Although the current version already represents a qualitative leap, the company envisions future iterations capable of modeling even more sophisticated macromolecular complexes, such as ribosomes and transcription complexes.

#deepmind#alphafold3#ia#proteínas#descoberta de medicamentos