Optimize proteins faster
Your Workflow with Exazyme
Using Exazymes AI, you can optimize proteins with up to 16 times less wetlab work. You provide us with initial measurements about the properties of some protein variants. We train an AI-model for predicting properties of variants you have not synthesized. Now we can automatically provide you with promising sequences. This way, your next round of wetlab experiments will be much better informed.
Upload your measurement data, i.e. variant sequences and measured properties.
Our models are fine tuned with your data to provide high prediction accuracy for properties the specific protein family.
You receive an email with some promising sequences that would likely improve the proteins properties or provide information about its fitness.
You synthesize and test the suggested sequences, thereby producing more measurement data. If you want to improve the protein even further, simply repeat the process.
As an EXIST project at the FU Berlin, Exazyme is funded by the BMWi and the European Social Fund, among others.