Artificial intelligence simplifies enzyme engineering
It is impossible to move a pharmacologist from a laboratory to a kitchen, and expect them to produce the same results. The same is true for enzymes: they are dependent on a particular environment. Researchers from Osaka University, however, have recently achieved this goal in a recent study published in ACS Synthetic Biology.
Enzymes are able to perform amazing functions because of the unique arrangement in which their amino acids are arranged. However, this is usually limited to a particular cellular environment. If you alter the cellular environment the enzyme will not function well, if at all. In order to improve the performance of enzymes under different conditions, such as those that favor biofuel production, it has been a research goal for many years. Such work traditionally involved extensive experimentation and trial-and error, which may have had little guarantee of an optimal outcome.
Artificial intelligence can reduce this trial and error, but it still relies upon experimentally determined crystal structures of enzymes — which can be unavailable or are not particularly useful. Teppei Niide is a co-senior researcher and says that \”the amino acids to mutate within the enzyme may only be best guesses.\” To solve this problem we developed a method of ranking amino acids based solely on the amino acid sequences of similar enzymes in other species.