Enzyme promiscuity is the ability of the enzyme to specifically catalyze certain reactions left to catalyze many other reactions. In some cases, promiscuity is associated with conformational diversity; that is, the conformational changes that have occurred in evolution allow the same enzyme to adapt to different substrates. This provides theretical support for the idea that the same enzyme may catalyze similar reactions.

Our algorithms to compute reaction similarity are explained in the article[1] introducing RxnSim. The main idea is to extract feature vectors from reactions or molecules and compute based on them. And the article also shows the ROC curve of the algorithm.

MEI chemical structure editor is based on open-source sofware Ketcher

Search algorithm of MEI is mainly based on tracking the change of the main group from substrate to product using the lasso button.

When switching from reactant to product or vice versa, the curren workspace would be temporarily saved. Choose one reaction type. Input the organism of enzyme will improve the performance.

The user could select cofactor, optimal pH and temperature and then begins the second search. MEI will return 4 kinetic constants: km1, km2, kcat1, kcat2. And they will be given as a range of values with the format of "[max,min]", as is shown below.

As we continue to refine our project, we do extensive dry lab validation to test our project model and performance. Our dry lab validation is based on the dataset of the 22nd (2003) edition of the IUBMB-Nicholson Metabolic Pathways Chart which contains updated pathways involved in ATP metabolism in the mitochondria and chloroplast. We selected multiple different types of reactions from this pathway to validate our software model, and we did get desirable results confirming the functional reliability of our software.

These are the validation result we get and we do get expected result which matches the above Metabolic Pathways' enzyme.

Also we do validations by furthur filtering information to get more accurate result, which supports our dry lab validation.

USTC Team offered the pathway they desgned to synthesize Borneol. Those reactions were then input to our website and search for potential enzymes. Certain enzymes selected were exactly the same as those mannually screened by USTC Team. Besides, our website found some enzymes which may contribute for future research.

We also invited NNU Team to validate our software. They searched for enzymes on our platform and compared the results with the ones the used in practice. It turned out that our platform successfully returned the enzymes they used, which inspired us a lot.

We visited Anhui Lianchuang Biological Medicine Co., LTD and discussed with their researchers. They offered us their data information during their production process. The data were then input to our platform and the results were given back to the corporation. The enzymes we selected were quite similar to the ones they used.

[1] Varun Giri, Tadi Venkata Sivakumar, Kwang Myung Cho, Tae Yong Kim, Anirban Bhaduri. RxnSim: a tool to compare biochemical reactions.Bioinformatics, Volume 31, Issue 22, 15 November 2015, Pages 3712–3714, https://doi.org/10.1093/bioinformatics/btv416