Fig. 1: Overview of CataPro.
From: Robust enzyme discovery and engineering with deep learning using CataPro
a Establishment of ten-fold unbiased dataset for kcat, Km, and kcat/Km. The enzyme sequence similarity between each fold data is lower than 0.4. b Kinetic parameter (kcat or Km) prediction framework based on language model embeddings and molecular fingerprints. ckcat/Km prediction framework based on kcat and Km pre-trained models.
