A new study has revealed new forms of proteins that have never been seen before along with unexpected connections. Last year, Google DeepMind predicted the structure of all known proteins from organisms using AlphaFold [1].
This time researchers mined a database full of the structures of every known protein predicted using AlphaFold neural networks [1]. This database consists of over 200 million protein structures. They have found completely novel shapes and connections between the structures.
A team co-led by Steinegger and computational biologist Pedro Beltrao, at ETH Zurich in Switzerland has developed a tool that can easily compare structures in the database based on their similarities and shapes. This tool is based on a structure alignment clustering algorithm- Foldseek cluster—that can cluster hundreds of millions of structures [2]. It can cluster millions of structures quickly.
This led to the identification of a novel link, as stated by Steinegger, among different proteins. A protein that humans and other complex organisms use to trigger an immune attack and to detect viral DNA was found in a cluster with proteins from single-celled bacteria and archaea.
Another study done at the University of Basel in Switzerland and the SIB Swiss Institute of Bioinformatics reveals a novel protein shape [3]. The researchers called it a ‘Beta-flower’ as the structures contain hairpin turns which are found in a known protein shape called a Beta-barrel that resembles petals on a flower [4]. According to computational biologist Joana Pereira, proteins that contain Beta-flowers are distantly related to one another, but their function is unclear.
References
- Jumper, J., Evans, R., Pritzel, A., Green, T., Figurnov, M., Tunyasuvunakool, K., … & Hassabis, D. (2020). AlphaFold 2. Fourteenth Critical Assessment of Techniques for Protein Structure Prediction; DeepMind: London, UK.
- Barrio-Hernandez, I., Yeo, J., Jänes, J. et al., (2023). Clustering-predicted structures at the scale of the known protein universe. Nature.
- Durairaj, J., Waterhouse, A.M., Mets, T. et al. (2023). Uncovering new families and folds in the natural protein universe. Nature.
- https://www.nature.com/articles/d41586-023-02892-z#ref-CR1