Today, SyntheticGestalt has announced the discovery of four new PETases (PET hydrolases) in a joint research project with the Tokyo Institute of Technology (hereinafter "Tokyo Tech"). This discovery is the result of using SyntheticGestalt's AI platform for discovering novel enzymes.
The project has been using the AI platform to discover PETases that are novel (less than 75% amino acid sequence homology to known enzymes) from a data set of approximately 230 million amino acid sequences of unknown function. As a result, four novel PETases were successfully discovered.
PETases are thought to have arisen as a result of the adaptive evolution of micro-organisms to wide presence of PET in nature as a result of mass production by humans. Therefore, relatively few micro-organisms with PETases have been discovered, making the discovery of novel enzymes under this constraint a challenging task.
Specifically, we discovered new enzymes from both PETases (Type 1 and Type 2), which are thought to be evolutionarily distinct lineages, and confirmed in collaboration with Associate Professor Kousuke Fujishima of the Tokyo Institute of Technology's Earth-Life Science Institute (ELSI) that they have substrate specificity for PET polymers with different physical shapes and crystallinity. These properties suggest that PETases may allow PET waste (e.g. food packaging), which is difficult to recycle with existing technologies, to be recycled in the future.
PET degradation technology by PETase enables low-cost, low-carbon and high-purity recycling compared to current recycling technologies (e.g. chemical and mechanical methods) and is expected to contribute to the realisation of a sustainable society.
SyntheticGestalt is an artificial intelligence research and development company specialising in AI drug discovery and other life science fields. Its research and development is not limited to drug discovery, but also focuses on the discovery of useful substances using its independently developed artificial intelligence technology. The AI platform used in this research has a cloud-based, scalable structure and can make predictions on large amino acid sequence libraries of several hundred million sequences, making it possible to predict not only PETases but also other various enzyme functions. SyntheticGestalt welcomes open innovation through joint research with public institutions and private companies.